Pollen biology and biotechnology 9781578082414, 9781138407732

419 41 10MB

English Pages 301 [316] Year 2003

Report DMCA / Copyright


Polecaj historie

Pollen biology and biotechnology
 9781578082414, 9781138407732

Table of contents :
Content: Pollen development --
Pollen morphology and aeropalynology --
Pollen viability and vigour --
In-vitro pollen germination and pollen tube growth --
Pollen sterility --
Pistil --
Pollination --
Pollen-pistil interaction and fertilization --
Self-incompatibility --
Interspecific incompatibility --
Optimization of crop yield --
Commercial production of hybrid seeds --
Transfer of useful genes to crop species --
Induction of haploids from pollen grains --
Production of other economic products.

Citation preview

Pollen Biology and Biotechnology

Pollen Biology and Biotechnology K.R. Shivanna

CRC Press

Taylor & Francis Group ^ ---- S

Boca Raton London New York

CRC Press is an imprint of the Taylor & Francis Group, an informa business


CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 First issued in hardback 2019 © 2003 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works ISBN-13: 978-1-57808-241-4 (pbk) ISBN-13: 978-1-138-40773-2 (hbk) This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint. Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www. copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe.

Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com Library of Congress Cataloging-in-Publication Data Shivanna, K.R. P o llen b i o l o g y a n d b i o t e c h n o l o g y / K .R . Shivanna. p . cm. Includes b i b l i o g r a p h i c a l r e f e r e n c e s (p. ). ISBN 1-57 8 0 8 - 2 4 1 - 2 1, Pollen. 2. P o l l e n - B i o t e c h n o l o g y . I. Title. QK6 5 8 . S 5 3 5 3 200-3 571.8'45--dc21


Dedicated to my Guru and Mentor Professor NS Rangaswamy

Preface The last three decades have been the most ex­ citing period in the history of polien biology. In­ tegrated studies using diverse techniques par­ ticularly of advanced microscopy, cell and mo­ lecular biology, and genetics have revolutionized our knowledge of the structural and functional aspects of pollen, leading to the development of a ‘new biology’ of pollen. This remarkable progress on pollen biology has resulted in the development of a number of effective and viable pollen-based technologies, broadly termed pol­ len biotechnology, for practical benefit, especially for crop improvement. The vast knowledge generated on pollen bi­ ology and biotechnology is scattered in Jour­ nals devoted to a number of disciplines such as morphology, physiology, biochemistry, cell and molecular biology, plant tissue culture, agricul­ ture, horticulture, forestry and plant breeding. The books available at present are confined to the coverage of a few selected areas of pollen biology/biotechnology. It has thus become diffi­ cult for teachers, students and researchers to get an overview of recent developments cover­ ing the whole spectrum of pollen biology and biotechnology. The primary objective of the present volume is to give a coherent and concise account of pollen biology and biotechnology with an em­ phasis on recent developments. The discussion is largely confined to pollen grains of flowering

plants. The number of papers published on pol­ len is so large that the over 1500 references cited had necessarily to be selective and subjective. The reader is able to access most of the litera­ ture in the field by referring to the cited papers and reviews. The topics included in the book are interdisciplinary and cater not only to pollen bi­ ologists and biotechnologists but also to repro­ ductive biologists, pollination biologists, aerobiologists, plant breeders, horticulturists and foresters. I hope that the book, apart from pro­ viding an overview of pollen biology and biotech­ nology to students, teachers and researchers will facilitate integration of pollen biotechnology into the traditional methods of crop production and improvement. My involvement with pollen goes back to 1964 when I started my Ph.D. programme under the supervision of Professor NS Rangaswamy. Another dimension was added to my career on pollen when I began, in 1974, collaborative research with late Prof. J Heslop-Harrison FRS, and Dr (Mrs) Y Heslop-Harrison, first at the Royal Botanic Gardens, Kew and then at the Welsh Plant Breeding Station, UK, which has continued over the years. My long association with Professor Rangaswamy and Heslop-Harrisons has immensely benefited my comprehension of pollen; I owe all those a great debt. I also had the privilege of discussing various aspects of pollen during my visits abroad with many


Pollen Biology and Biotechnology

renowned pollen biologists and biotechnologists, in particular Late Prof. RB Knox, Prof. HF Linskens, Prof. M Cresti and Prof. VK Sawhney. I thank them most sincerely for their time, interest and encouragement. It is a pleasure to acknowledge the help and encouragement received over the years from many of my senior colleagues in the Department in particular Prof. BM Johri, Prof. HY Mohan Ram, Prof. PS Ganapathy and Prof CR Babu. I have greatly benefited in my career from the

April 2002

researches of and discussions with my students on pollen biology and biotechnology for over 25 years. Also, Dr Madhu Bajaj and Dr Rajesh Tandon, my former students, read the entire text of this book and gave constructive suggestions. I convey my deepest feelings of gratitude to all my students. I thank Mr Bhaskar Bhandari for his very able help in preparation of illustrations. I cannot adequately thank my wife Pramila for her understanding and encouragement throughout my career.

KR SHIVANNA Department of Botany University of Delhi

Contents Preface

vii Part 1: Pollen Biology

1. Pollen Development Differentiation of Anther Layers Microsporogenesis Initiation of Meiosis Synthesis of Macromolecules Cytoplasmic Reorganization Syncytium and Isolation Significance of Cytoplasmic Reorganization and Isolation Microgametogenesis Vegetative and Generative Cells Sperm Cells Dimorphism of sperm cells and organization ofmale germ unit Tapetum Secretory Tapetum Plasmodial Tapetum Tapetal Membrane Role of Tapetum Supply of nutrients to developing pollen Breakdown of callose wall around microspore tetrads Supply of sporopollenin precursors to pollen exine Supply of pollen coat substances and exine proteins Gene Expression during Anther Development Transmission of Pathogens through Pollen Pollen Maturation and Anther Dehiscence

2. Pollen Morphology and Aeropalynology Pollen Morphology Applications

7 7

8 8 9

10 10 12 13 14 16 17 18

20 20

21 21 21 21

22 22

22 23 23

26 26 26

x Pollen Biology and Biotechnology

Petroleum exploration Archaeology Criminology Testing purity of honey Three Domains of Pollen Wall Intine Exine Pollen coat substances Pollen Wall Morphogenesis Control of exine pattern Pollen wall proteins Pollen Analysis Pollen size and shape Compound pollen Polarity Apertures Exine sculpture LO analysis Aeropalynology Allergic Response Diagnostic Tests Allergens Biological Standardization of Allergens

3. Pollen Viability and Vigour Pollen Viability Tests for Viability Fruit- and seed-set Pollen germination and pollen tube growth in the pistil Non-vital stains and other tests of limited use Tétrazolium test In-vitro germination test Fluorescein diacetate test Causes for Loss of Viability Loss of membrane integrity Quantitative changes in phospholipids Phase transition and membrane integrity Pollen Vigour Tests for Vigour In-vitro germination Semivivo technique In-vivo pollen germination and pollen tube growth Effects of Environmental Stresses on Pollen Quality

4. In-vitro Pollen Germination and Pollen Tube Growth Germination Requirements Hydration

26 26 26 27 27 27 27 29 29 31 33 33 33 36 37 38 39 39 40 41 43 43 44

45 45 46 46 47 47 47 49 49 50 51 53 53 56 56 56 57 57 58 61 61 61

Contents xi

Carbohydrate Source Boron Calcium Effects of other Physical and Chemical Factors Phases of Germination and Tube Growth Lag Phase Pollen Tube Emergence Pollen Tube Growth Role of cytoskeleton and calcium in tip growth In-vitro vs in-vivo Tube Growth Release of Metabolites RNA and Protein Synthesis In-vitro Pollen Germination Assay to Study the Effects of Toxic Chemicals

5. Pollen Sterility Genic Male Sterility Phenotypic Effects of GMS Structural and Biochemical Changes Hormonal Changes Effects of Temperature and Photoperiod Induction of Pollen Sterility through Recombinant DNA Technology Cytoplasmic Male Sterility Alterations in Mitochondrial Genome Structural and Biochemical Changes Mechanism of Cytoplasmic Male Sterility Environmentally Induced Pollen Sterility Chemically Induced Pollen Sterility

6. Pistil Stigma Style Ovule and Embryo Sac

7. Pollination Breeding Systems Outbreeding Devices Secondary Pollen Presentation Pollen:Ovule Ratio Modes of Pollination Anemophily Hydrophily Zoophily Evolutionary Significance of Insect Pollination Floral Attractants and Rewards Floral Thermogenicity Pollen Travel and Gene Flow Pollination Postulates

62 64 65

66 67 67

68 69 70 72 73 73 76

77 77 77 78 80 80 80 82 84 85 87 89 89

90 90 93 98

102 104 104 104 105 105 105 105 106 109 110

111 111 113


Pollen Biology and Biotechnology

Pollination Efficiency

8. Pollen-Pistil Interaction and Fertilization Significance of pollen-Pistii Interaction Screening for Compatibility Screening for Quality Pollen Viability and Stigma Receptivity Events on Stigma Surface Pollen Adhesion and Hydration Pollen Germination and Pollen Tube Entry into Stigma Pollination Stimulus Pollen Tube Growth through Style Regulation of Pollen Tube Number Internal Geitonogamy Pollen Tube Guidance On Stigma Surface In Ovary Pollen Tube Entry into Ovule and Embryo Sac Double Fertilization Preferential Fertilization Inheritance of Plastids Positional Effect of Ovules In-vitro Pollination and Fertilization In-vitro Pollination Pollination of cultured pistils Pollination of cultured ovules In-vitro Fertilization Applications of in-vitro Pollination and Fertilization

9. Self-incompatibility Evolution of SI Homomorphic SI Genetics Cytology Physiological and Biochemical Studies Temporal Expression of SI Operation of SI Characterization of S-allele Products and S-alleles in Pistil Sporophytic systems Gametophytic systems Papaver system S-allele-specific components in pollen Basis of S-allele Specificity Mechanism of Pollen Recognition and Inhibition Heteromorphic SI Dimorphic Systems Trimorphic Systems


114 114 115 115 117 118 118 119 120 121 124 125 125 127 128 130 131 131 132 133 133 134 134 134 137 138 140 140 141 141 144 146 148 149 149 149 152 153 153 154 154 158 158 163

Contents xiii

Zone of Inhibition Mechanism of Inhibition Passive inhibition Combination of passive and active inhibition

10. Interspecific Incompatibility Prefertilization Barriers Unilateral Incompatibility Incongruity/Passive Inhibition Mechanism of Passive Inhibition Post-fertilization Barriers

163 165 165 166

167 167 167 168 170 175

Part II: Pollen Biotechnology 11. Optimization of Crop Yield Enhancing Populations of Native Pollinators through Habitat Management Use of Commercially Managed Pollinators Spraying Pollinator Attractants on Target Crop Introduction of Pollinators Supplementary Pollination/Assisted Pollination

12. Commercial Production of Hybrid Seeds Use of Genic Male Sterility Vegetative/Micropropagation of Female Lines Use of Phenotypic Markers Environmental/Hormonal Induction of Male Fertility Use of Cytoplasmic Male Sterility Use of Self-incompatibility Methods to Overcome SI Selection of Lines with Strong SI Alleles Transfer of SI Alleles to Self-compatible Species Hybrid Seed Production in Monoecious Species Manipulation of Sex-expression for Yield Improvement Use of Pollen Sterility Induced through r-DNA Technology Use of Pollen Sterility Induced through Chemical Hybridizing Agents Proline Analogues Phenyl Pyridazones Phenylcinnoline Carboxylates LY 195259 MON 21200 Potential of Apomixis in Hybrid Seed Technology

13. Transfer of Useful Genes to Crop Species Use of Pollen for Identifying Plants with Desirable Genes Production of Hybrids Effective Techniques to Overcome Physical Barriers Pollen storage Effective Techniques to Overcome Prefertilization Barriers

181 181 182 183 184 184

186 187 188 189 189 189 190 191 193 193 193 194 194 194 195 195 196 196 196 196 199 200 202 203 204 206

xiv Pollen Biology and Biotechnology

Application of growth substances and other chemicals Use of mentor pollen Stump pollination Stylar grafting Bridge crosses Bud pollination Intraovarian pollination In-vitro pollination of ovules In-vitro fertilization Methods to Overcome Post-fertilization Barriers Multiplication of Hybrids Identification of Hybrids at Seedling Stage Induction of Amphiploidy Handling of Backcross Generations Application of selection pressure to pollen Stabilization of Recombinants Use of Pollen for Genetic Transformation Mature Pollen Grains Microspores and Microspore-derived Embryos

14. Induction of Haploids from Pollen Grains Importance of Haploids Production of Homozygous Diploids Recovery of Novel Recombinants Mutation Research Model System to Study Embryogenesis Production of Artificial Seeds and Genetic Transformation Production of Haploids Use of Genetic Lines Chromosome Elimination Culture of Anthers and Microspores Optimal Conditions for Induction of Pollen Embryos Pollen dimorphism Diploidization Developmental Pathways Production of Haploids through Gynogenesis Utilization of Haploids

15. Production of Other Economic Products Pollen as Health Food Supplement Pollen as Medicine Pollen Grains/Pollen Allergens for Diagnosis and Therapy Treatment of Pollen Allergy Pollen Calendars

207 208 208 209 209 210 210 210 210 211 211 212 212

212 213 215 215 216 217

219 219 219 219 220 220 220 220 220 221 222 223 225 225 225 227 229

231 231 231 233 233 234






The role of pollen grain as the male partner in sexual reproduction of seed plants was estab­ lished by the end of the 19th century (Maheshwari 1950,1963). Pollen grains develop in anthers as a result of reduction division and represent highly reduced male gametophytes. In gymnosperms the microspore undergoes 3­ 5 mitotic divisions to produce male gametes. At the time of shedding, the gymnosperm pollen grains contain a variable number of prothallial cells, a tube cell and an antheridial cell; follow­ ing pollen germination the antheridial cell divides to form a stalk cell and a body cell. The body cell eventually divides and produces two male gametes (Pacini 2000). In flowering plants the microspore undergoes only two mitotic divisions to produce male gametes. At the time of shed­ ding pollen grains of flowering plants contain either 2 cells (vegetative and generative) or 3 cells (vegetative cell and two male gametes (formed by division of the generative cell)) at the time of dispersal (Fig. 1A-C). Following anther dehiscence, pollen grains are released into the atmosphere. At the time of release, they are desiccated and their moisture level is generally reduced to less than 15%. Pol­ len grains survive for varying periods after re­ lease as independent functional units. This phase of pollen, termed the free dispersed phase, is an important aspect of pollen biology as it facilitates gene flow. In gymnosperms, pollen grains land at the micropylar part of the ovule that houses the fe­ male gametes inside the archegonia (Fig. 1 D). A pollination drop is produced at the micropyle in most of the gymnosperms and aids in pollen passage into the ovule and in pollen germina­ tion. The male gametes are released near the archegonia, pass through the neck of the arche­ gonia, reach the egg cell and effect fertilization (Friedman 1993). In flowering plants, pollen grains are deposited on the stigma (Fig. 1 E) where they germinate and the resultant pollen tubes grow through the tissues of the stigma and style, enter the ovule and eventually the embryo sac (female gametophyte) present inside the

ovule, where the two male gametes are dis­ charged. This process of conveying the male gametes through directional growth of the pol­ len tube is termed siphonogamy. Apart from flow­ ering plants, siphonogamy is also present in advanced gymnosperms such as Pinus and Gnetum. This phase of pollen biology, from pollination until the pollen tube reaches the embryo sac, is generally referred to as pollen-pistil interaction. Pollen-pistil interaction plays an important role in sexual reproduction. The pollen grain is screened during pollen-pistil interaction; the pistil facilitates germination of compatible pollen and growth of the resultant pollen tube until it reaches the embryo sac while incompatible pollen is in­ hibited before germination or during the growth of pollen tube before its entry into the embryo sac. One of the two sperms released in the em­ bryo sac fuses with the egg to form the zygote and the other with the secondary nucleus of the central cell to form the primary endosperm nucleus, thus completing the process of double fertilization, characteristic of flowering plants.The zygote gives rise to the embryo and the primary endosperm nucleus to the endosperm which nourishes the embryo. The ovule develops into the seed and the ovary into the fruit. Thus, de­ velopment of functional pollen, their transfer to the stigma and successful completion of pollenpistil interaction and fertilization are the prereq­ uisites for fruit- and seed-set. Pollen biology in­ volves a comprehensive understanding of the structural and functional details associated with the aforesaid pollen events. Seeds and fruits are the economic products of more than 90% of our crop plants. A thorough knowledge of pollen biology and its manipula­ tion are required for any rational approach to increase crop productivity.This realization, in re­ cent years, has greatly stimulated researches on pollen biology. Early studies on pollen biol­ ogy used only conventional techniques such as microtomy and acetolysis. Since the 1950s, a number of physiological and biochemical tech­ niques were introduced particularly in studies

4 Pollen Biology and Biotechnology


generative £ cell

- vegetative ceil vegetativej I nucleus

tube cell

_ generative | cell sperm cells ■


stigma micropyle integument


pollen chamber archegonial chamber archegonium egg cell

transmitting tissue

nucellus female gametophyte placenta ovule ovary embryo sac

Fig 1- Male (A-C) and female (D, E) gametophytes in seed plants. A, D. Gymnosperms, B, C and E. Angiosperms.

on pollen viability, storage and germination. In­ vention of electron microscopy in the 1960s added another dimension to the study of pollen. Detailed fine structural studies have since been carried out on almost all facets of pollen biology.

In recent years pollen biologists have been us­ ing better fixatives, more diverse and versatile microscopes, and the techniques of cell and tis­ sue culture, and immunofluorescence. During the last 15 years techniques of molecular

Pollen Development 5

biology and genetics have been integrated effectively in studies on pollen biology. These integrated approaches have revolutionized our knowledge of the structural and functional as­ pects of pollen. The demonstration that pollen grains of many species present in air are responsible for allergy led to comprehensive studies on the distribu­ tion of pollen in air and the details of allergy. This area of pollen biology has now developed into an independent branch, aeropalynology (Knox 1979, Mohapatra and Knox 1996). Since pollen morphology is a conserved character and often specific to a given species/ genus, it has been used widely in understand­ ing taxonomic and phylogenetic relationships of species. Studies on fossil pollen grains have added greatly to our understanding of the ori­ gin, distribution and evolution of flowering plants. Also, because of the correlation that exists be­ tween pollen flora in geologic ages and the pres­ ence of petroleum fuel, studies on fossil pollen have become an integral part of the petroleum industry (Faegri and Iversen 1989). Pollination is another area that has long attracted the attention of pollen biologists (McGregor 1976, Faegri and van der Pijl 1979, Real 1983, Free 1993). Initial studies were largely confined to identifying pollinating agents and floral adaptations. Gradually the pollinatorflower interaction and co-evolution of the flower and the pollinator have become the main focus. In-depth studies are being carried out on pollinator attractants and rewards, specialization of the flower for pollinators and the breeding system (Richards 1986). Two manuals on the techniques of pollination biology were recently published (Dafni 1992, Kearns and Inouye 1993). Pollen-pistil interaction was largely ignored until the 1960s. Following the realization that the main barrier to fertilization occurs during pollenpistil interaction, interest in studies on pollenpistil interaction has steadily increased over the years and this aspect has become one of the

front-line areas of research during the last two decades (Raghavan 1999, de Nettancourt 2001). Fertilization in flowering plants has been a difficult area for research as it takes place deep inside the ovule and imposes technical problems for basic studies as well as effective experimen­ tation. One of the important goals of embryolo­ gists has been to achieve in-vitro fertilization using isolated egg and sperm cells so as to fa­ cilitate basic as well as experimental studies on fertilization in flowering plants. Considerable advances have been made in recent years on the structural details of fertilization at the light as well as electron microscopic levels. In-vitro fertilization using isolated egg and sperm cells has now become a reality and progress in un­ derstanding fertilization is expected to be rapid in the coming years (Kranz and Dresselhaus 1996, Shivanna and Rangaswamy 2000). Studies on pollen biology have also acquired significance in the light of recent progress in de­ veloping transgenic plants in several crop spe­ cies and their release. Since engineered genes from crop species can escape and spread to other species or cultivars through pollen, gen­ eration of detailed information on the extent of pollen flow, pollen viability and its compatibility with related species has become a requirement for the release of transgenics (Bhatia and Mitra 1998). Extensive literature accumulated on various aspects of pollen biology has been reviewed in a number of books over the years (Maheshwari 1950, 1963, Stanley and Linskens 1974, de Nettancourt 1977,2001, Knox 1979, Johri 1984, Shivanna and Johri 1985, Raghavan 1999). Many edited volumes have also appeared on different areas of pollen biology/biotechnology (Mulcahy et al. 1986, Linskens 1974, Mulcahy and Ottaviano 1983, Cresti et al. 1988, 1992, Ottaviano et al. 1992, Russel and Dumas 1992, Williams et al. 1994, Mohapatra and Knox 1996, Shivanna and Sawhney 1997, Clement et al. 1999, Dafni et al. 2000).

1 Pollen Development In flowering plants pollen grains develop in the anther. A young anther consists of a homoge­ neous mass of meristematic cells surrounded by an epidermis. Further development of the anther involves a major histodifferentiation re­ sulting in several highly specialized cells and tis­ sues. Some of these cell types continue to dif­ ferentiate while others degenerate. The differ­ entiation and degeneration events take place in a precise spatial and temporal order in the an­ ther and ultimately result in the development and dispersal of pollen grains (Goldberg et al. 1993). Details of pollen development are quite uniform in different species and have been reviewed from time to time (Heslop-Harrison 1971, 1972, Mascarenhas 1975,1989, Shivanna and Johri 1985, Raghavan 1999). The following are the major events associated with pollen develop­ ment.

DIFFERENTIATION OF ANTHER LAYERS The mature anther is usually a bithecous and tetrasporangiate structure. Different cell and tis­ sue layers demarcated in the mature anther can be traced to the three primary germ layers—Lv L2 and l_3—present in the floral meristems (see Goldberg et al. 1993). L, gives rise to the epi­ dermis, L, (hypodermal layer) to the archesporium and L3 to the connective tissue and the

central vascular bundle (Fig 1.1 ).The epidermis undergoes no further elaboration except for dif­ ferentiation of the stomium, along the line of anther dehiscence in the mature anther.The con­ nective tissue also shows only a limited differ­ entiation; it gives rise to the inner tapetum which lines the inner portion of the anther locules and a small cluster of cells separating the two pollen sacs of each half anther (intersporangial sep­ tum/circular cell cluster) along the stomium. Histodifferentiation events of the anther are largely confined to the derivatives of the hypo­ dermal layer (L2). As the anther develops, it as­ sumes a four-lobed appearance and four groups of archesporial cells, corresponding to the four microsporangia (one in each anther lobe), dif­ ferentiate in the hypodermal position. The archesporial cells are easily distinguishable from other cells by their dense cytoplasm and large nuclei. Archesporial cells divide periclinally to form an outer primary parietal layer and an in­ ner primary sporogenous layer. The former un­ dergoes a few more periclinal divisions and gives rise to the endothecium, 2 or 3 middle layers and the outer tapetum lining the outer face of the anther locules. The primary sporogenous layer undergoes a few mitotic divisions and gives rise to the microspore mother cells. The microspore mother cells and the tapetum are the sites of intense metabolic activity during meiosis.

8 Pollen Biology and Biotechnology

Anther primordium


L1 I

L3 J_

archesporial cells



______ I______

connective vascular bundles

primary parietal cells

sporogenous cells



secondary parietal cells — endothecium



L2 I


— middle layers

vascular bundles

microspore mother cells


outer tapetum


microspore tetrads



É connective inner tapetum senesce


splits senesce longitudinally



pollen grains

▼ degenerate



Fig. 1.1 Schematic representation of cell lineages from the three floral meristem layers—L1, L2, L3 (modified from Goldberg et al. 1993)

MICROSPOROGENESIS Meiosis is one of the most critical events during microsporogenesis and thus in the development of pollen. Apart from reduction in number of chro­ mosomes, it facilitates genetic recombination, the most significant feature of sexual reproduc­ tion. Anthers provide an excellent system for studying meiosis because: (a) they are easily available and accessible for experimental ma­ nipulation, (b) each anther contains a large num­ ber of meiocytes, more or less in synchrony, and (c) different anthers of a flower bud often show good synchrony so that one of them can be used for identifying the stage of meiosis while others can be used for experimentation. Anthers of Lilium and Trillium are the most widely used in these studies because of their large size and good synchrony. The duration of meiosis is highly variable in different species; it is o • 0(

°X w ,,

0 • \ V it v'cP \ *

So *-% O

o o0 /






° O

c 0


0 o/



I . ° , A _so 4 V » **

generative cell

— vegetative — nucleus generative cell -

Fig. 1.4 Diagrammatic representation of asymmetric division of the microspore. A. Microspore soon after release. 6 -D. Formation of vacuole (v), migration of the nucleus, polarized distribution of mitochondria and plastids away from the nucleus, and division of microspore. E-H. Separation of generative cell from the microspore wall and resorption of vacuole.

Pollen Development 15

long axis (Cresti et al. 1984, Derksen et al. 1985, Pierson and Cresti 1992). The GC shows nor­ mal profiles of endoplasmic reticulum (ER), ri­ bosomes, dictyosomes and mitochondria. The cytoplasm of the VC contains all the nor­ mal organelles (Sanger and Jackson 1971a, b). The number of organelles increases steadily until pollen maturation. The nucleus of the VC is generally lobed in mature pollen. The plastids, which are free from starch grains at the vacu­ olate microspore stage, develop starch grains soon after the division. Reserve food material in the form of lipids or starch grains accumulates in the VC; this is generally associated with the breakdown of the tapetum (Echlin 1972, Christensen and Horner 1974), indicating that the degradation products of the tapetum are taken up by the developing pollen. In many taxa, such as tobacco, stacked rough endoplasmic reticulum (RER) appear in the vegetative cell (Jensen et al. 1974, Cresti et al. 1975). Although both the VC and the GC are prod­ ucts of mitotic division of the microspore, the two cells are structurally and functionally very differ­ ent. Many differences have been recorded in the chromatin of the two nuclei (Table 1.1). These differences are considered responsible for the differences observed in the transcriptional and translational activities of the two cells (Reynolds and Raghavan 1982) Protocols for isolation of a large number of generative nuclei free from vegetative nuclei Table 1.1 Differences between vegetative cell and genera­ tive cell Feature

Vegetative ceil

Generative cell


Active with high

Almost dormant

DNA Nucleus


levels of RNA and

except for a mitotic

protein synthesis

division in 3-celled pollen Increases to 2C level

Remains at 1C level Low amount of Higher amount of DNA-associated DNA-associated lysinelysine-rich histone rich histone Diffuse Highly condensed

have been developed (Ueda and Tanaka 1994). Isolated generative nuclei have been used to characterize basic protein components (Ueda and Tanaka 1994, 1995). These studies have shown the presence of five basic proteins spe­ cific to or highly concentrated in the generative nuclei with apparent molecular masses ranging from 18.5 to 33 kD. Further analyses of two of these proteins (18.5 and 22.5 kD) revealed that although they resembled somatic histones H2B and H3 respectively, histones from the genera­ tive nucleus showed unique peptide fragments and their antibodies did not cross-react with any somatic histones. One of these proteins (22.5 kD) is lysine-rich and the other (18.5 kD) argin­ ine-rich. Immunofluorescence studies showed that both proteins are present only in the gen­ erative nucleus and not in the vegetative. The technique of genetic ablation which in­ volves targeted expression of a cytotoxic pro­ tein under the control of cell-specific regulatory sequences has been used to understand the ex­ pression of some genes in the two cells and their functional interaction. In tobacco, the promoter of the LAT 52 gene, which is activated specifi­ cally in the VC following pollen mitosis, was used to express cytotoxic diphtheria toxin A chain (DTA) (Twell 1994). Analysis of pollen of such transgenic plants showed that LAT 52-DTA re­ sulted in ablation of VC soon after microspore division (as revealed through FDA test).The GC retained its viability for several days following VC ablation but progressively lost viability. Interest­ ingly, absence of functional VC prevented migration of the GC away from the pollen grain wall into the VC cytoplasm. These studies el­ egantly demonstrate the dependence of GC on the VC for its functioning. Several studies have shown that the vegeta­ tive nucleus forms an intimate association with the GC in mature pollen or during pollen tube growth (Ciampolini etal. 1988, Hu and Yu 1988, Wagner and Mogensen 1988). In Medicagosativa (Shi et al. 1991) the vegetative nucleus in mature pollen surrounds a major part of the GC. The nuclear pore density on the surface of the

16 Pollen Biology and Biotechnology

vegetative nucleus facing the GC is 69% higher than that on the surface away from the GC. As the macromolecular traffic into and out of the nucleus occurs through the pores, these obser­ vations suggest a functional relationship be­ tween the vegetative nucleus and the GC. This is in agreement with the report (Tang 1988) on the increased level of ATPase activity at the veg­ etative nuclear envelope in the area of its asso­ ciation with the GC in Amaryllis and Clivia. An interesting variation in the disposition of the GC has been reported in Rhododendron (Kaul at al. 1987) and Acacia (McCoy and Knox 1988), both characterized by polyads. In Rhodo­ dendron the spindle-shaped generative cell possess two extensions, one at either end. One of these extensions is connected to the intine wall ingrowth situated near the external aper­ ture (Theunis et al. 1985). Following in-vitro ger­ mination, this connection breaks off and the gen­ erative cell enters the pollen tube where it be­ comes closely associated with the vegetative

nucleus (Kaul et al. 1987). In Acacia (McCoy and Knox 1988) the spindle-shaped GC lies across the width of the pollen grain and is attached to both ends of the intine by means of membrane labyrinths.

Sperm Cells The GC divides mitotically to form two sperm cells (Fig. 1.5 A, B). In species in which pollen grains are shed at the 2-celled stage, the GC generally enters dormancy at the prophase stage. In species in which pollen grains are shed at the 3-celled stage, the GC completes division before dormancy. Early electron microscopic studies on the division of the GC in barley (Cass and Karas 1975) indicated that the nuclear divi­ sion of the GC takes place while it is still attached to the pollen wall. An incipient wall appears be­ tween the two sperm cells and the two-celled unit detaches from the pollen wall. Eventually the partition wall as well as the surrounding walls of the sperm becomes degraded, releasing the

• 7 “ %

— $

1 \


i - ' vn



Fig. 1.5 Fluorescence micrographs of squash preparations of pollen grain (A), and pollen tube (B) of tobacco stained with DAPI, a DNA fluorochrome, to show the vegetative nucleus (vn), the generative nucleus (gn) and the sperm nuclei (sn). In B pollen tube tip is toward the top (after Shivanna and Rangaswamy 1992).

Pollen Development 17

sperms. Subsequent studies on barley (Charzynska et al. 1988) and a few other spe­ cies (Charzynska et al. 1989, Murgia et al. 1991) have shown that the GC detaches from the pol­ len wall before its division and a regular cell plate is formed between the two nuclei during divi­ sion. In several species the two sperm cells re­ main associated with each other even after divi­ sion as part of the male germ unit (see below) In nearly 70% of the flowering plants, pollen grains are shed at the 2-celled stage (Fig 1.5A). In the remaining species they are shed at the 3celled stage (Brewbaker 1959). In 2-celled pol­ len the generative cell divides and gives rise to the two sperm cells in the pollen tube after ger­ mination (Fig.1,5B).The cytological status of the pollen (2 or 3 cells) at the time of shedding is correlated with a number of physiological and genetical features (Table 1.2). Table 1.2 Correlation between pollen cytology and some other traits Character

2-celled pollen

3-celted pollen

Viability Storage In-vitro germination

prolonged store well comparatively easy low gametophytic

short difficult to store comparatively difficult high sporophytic

Respiratory rate Self-incompatibility

Dimorphism of sperm cells and organization of male germ unit The classical concept assumes that the two sperm cells formed by the division of the GC are isomorphic. Recent studies in several spe­ cies have clearly shown distinct morphological differences between the two sperm cells of the pollen. Such studies became possible only after the technical advances made to produce com­ puter-assisted three-dimensional reconstruction of serial transmission electron micrographs. In Plumbago (Fig. 1.6 A) (Russell and Cass 1981, Russell 1984) the sperm cells are linked together by a common transverse wall traversed by plasmodesmata. One of the sperm cells (Svn) is

associated with the vegetative nucleus through a long extension. The two sperm cells also show differences in nuclear size and number of cytoplasmic organelles (Table 1.3) (Russell 1984). The smaller sperm, which is not associ­ ated with the vegetative nucleus (Sua), contains an average of 24 plastids and 40 mitochondria, while the larger sperm cell (Svn) usually con­ tains no plastids or very few plastids and has an average of 256 mitochondria. Such a physical association between the two sperm cells and the vegetative nucleus (linking all nuclear and cytoplasmic DNA of male heredity units) has been termed the ‘male germ unit’ (MGU). The MGU in Plumbago is formed in mature pollen and is maintained throughout pollen tube growth. Tablel .3 Quantitative details of the two sperm cells in Plum­ bago zeylanica studied through three-dimensional recon­ struction (based on data presented in Russell 1984) Feature Mean cell volume (pm3) Mean cell surface (pm2) Nuclear volume (pm3) Nuclear surface (pm2) Cytoplasmic volume (pm3) No. of mitochondria No. of plastids



69.5 147.9 19.9 32.4 45.2 256.2 0.45

48.9 84.7 12.1 20.4 33.4 39.8 24.3

Sperm dimorphism and MGU have subse­ quently been demonstrated in many other 3celled pollen species such as Beta vulgaris (Hoefert 1969), Catananche caerulea (Barnes and Blackmore 1987) and a few other taxa (Knox and Singh 1987, Mogensen 1992). In Brassica (Fig 1 .6B) the two sperm cells are linked by a common cell junction and one of the sperm cells is associated with the vegetative nucleus. Plas­ tids are absent in both sperm cells; however, the sperm cell associated with the vegetative nucleus (Svn) has a larger number of mitochon­ dria than the other sperm cell (Sua) (McConchie et al. 1985, 1987). Although differences in the contents have not been documented in other species, the difference in size of the two sperm cells has been reported in several species: e.g.

18 Pollen Biology and Biotechnology plastids


vegetative nucleus nucleus Sua microfi laments vegetative nucleus mitochondria

A Fig. 1.6 Diagrammatic representation of the male germ unit in Plumbago zeylanica (A) and Brassica campestris (B). Long projection of one of the sperms, Svn (sperm physically associated with vegetative nucleus) wraps around and lies within embayments of the vegetative nucleus in A. In B, Svn projection enters the enclaves within the vegetative nucleus. In Plum­ bago the Svn contains a majority of mitochondria and only two plastids whereas the Sua (sperm physically unassociated with vegetative nucleus) contains most of the plastids and fewer mitochondria. In Brassica the plastids are absent in both the sperm cells but the Svn is mitochondria-rich compared to Sua (A—after Russell 1984, B—Knox and Singh 1987).

Gladiolus and Rhododendron (Shivanna et al. 1987), Euphorbia and Gerbera (Mogensen 1992). The concept of MGU has now been extended to 2-celled pollen species in which the vegeta­ tive nucleus and the generative cell are in asso­ ciation (Knox and Singh 1987). Some of the 2celled pollen systems which show MGU are: Gossypium (Jensen and Fischer 1968), Hippeastrum vittatum (Mogensen 1992), Petu­ nia hybrida (Wagner and Mogensen 1988), Nicotiana tabacum (Yu et al. 1989), Rhododendron sp. (Shivanna et al. 1987), Acacia retinoids (McCoy and Knox 1988), Medlcago sativa (Zhu et al. 1992), Aloe ciliaris (Ciampolini et al. 1988) and Gladiolus gandavensis (Shivanna et al. 1987). Mature pollen grains of Poaceae, which are 3-celled, show neither sperm dimorphism nor MGU (Mogensen and Rusche 1985, Mogensen 1992). However, a close association is formed between the vegetative nucleus and the sperm cells soon after pollen germination, thus form­ ing the MGU (Mogensen and Wagner 1987).The association between vegetative nucleus and sperm cells is transitory and short; by the time sperm cells enter the pollen tube, the vegeta­

tive nucleus gets separated but the two sperm cells remain connected during pollen tube growth until the pollen tube approaches the ovule. A similar situation exists in Zea also (Rusche and Mogensen 1988). However, in Zea the vegetative nucleus remains associated with the sperm cells during their passage in the pol­ len tube. The significance of MGU in fertilization is not clearly understood. MGU, in particular the as­ sociation of the two sperm cells in the pollen tube, seems to be common. So far, analysis of sperm cells during pollen tube growth and its entry into the ovule has been confined to a lim­ ited number of taxa. MGU is likely to play an important role in the transfer and discharge of the two sperm cells together into the synergid (Russell 1992).

TAPETUM The tapetum is a transitory layer that surrounds the sporogenous tissue and plays an important role in pollen development (Pacini 1997). In most species the origin of the tapetum is dual. The outer tapetum (facing the outer surface of the anther) is derived from the secondary parietal

Pollen Development 19

layer, while the inner tapetum (facing the inner side of the anther locule) is derived from the cells of the connective tissue (Goldberg et al. 1993). In general, the tapetum is single layered. In some species, especially of marine angiosperms, the tapetum is 2-4 layered (Ducker et al. 1978). The tapetal cells are metabolically very ac­ tive during both the premeiotic and meiotic pe­ riod; they show active synthesis of RNA and pro­ teins. This is correlated with the presence of welldeveloped mitochondria, abundant RER cisternae and active dictyosomes in the cytoplasm of tapetal cells (Steer 1977). A unique cytological feature of tapetal cells is an increase in their DNA content initiated soon after the onset of meiosis in the sporogenous cells. As this increase in DNA is not followed by regular cell division, it results in one or more cytological abnormalities: multinucleate cells, polyploid nuclei (formed by in­ complete mitosis or nuclear fusion) and polyteny. Often the extent of DNA increase in the tapetal cells is up to 16 times that present in the sporogenous ceils (for a general review of the tapetum see Echlin 1971, Heslop-Harrison 1972, Bhandari 1984, Pacini et al. 1985, Chapman

1987, Pacini 1990, 1997). In members of Mimosoideae and Papilionoideae (Leguminosae), the tapetal cells remain uninucleate, a condition considered to be derived from bi- and multinucleate conditions (Buss and Lersten 1975, Albersten and Palmer 1979). Two main types of tapetum are distinguished: secretory/parietal tapetum and plasmodial/invasive/amoeboid tapetum.The distinction is largely based on the nature of the tapetal cells at the peak of their activity. In the secretory tapetum the cells of the tapetum maintain their position and identity and eventually undergo degenera­ tion in situ towards the end of pollen develop­ ment. In the plasmodiai tapetum the inner tan­ gential wall breaks down and the protoplasts of the tapetal cells enter the anther locule (Fig. 1.7A, B). The secretory tapetum is more com­ mon in dicots than in monocots. According to one survey (Pacini et al. 1985), the secretory tapetum occurred in 154 dicot and 21 monocot families, while a plasmodiai tapetum was present in 14 dicot and 18 monocot families. Both types of tapeta have been reported in 11 dicot and one monocot family.

tapetal protoplast

microspores tapetal — plasmodium



Fig 1.7 Diagrammatic representation of plasmodiai tapetum at microspore stage. Protoplasts of tapetal cells have entered anther locule between the microspores. The identity of individual protoplasts is maintained in A, whereas the protoplasts have fused to form tapetal plasmodium in B (after Shivanna et al. 1997).

20 Pollen Biology and Biotechnology

Secretory Tapetum Cells of the secretory tapetum retain plasmodesmatal connections between the neighbouring cells. In some species they develop cytomictic channels similar to those of sporog­ enous cells and thus form a tapetal syncytium (Heslop-Harrison 1972, Clement and Audran 1995). In several species, the inner tangential wall and later the radial walls break down, but the protoplasts remain in situ. Transfer of prod­ ucts from the tapetum is an integral part of pol­ len development. Transport of substances from the tapetal cells into the locule may occur through exocytosis (Ciampolini et al. 1993) or through transport across the inner plasma membrane. In Brassica oleracea the inner tapetal membrane forms tubular evaginations that increase the surface area and might be in­ volved in translocation of solutes (Murgia et al. 1991). The time of tapetum degeneration varies greatly from species to species. It is generally towards the end of pollen development but in Pterostylis, an orchid, tapetal degeneration oc­ curs before microspore mitosis (Fitzgerald et al. 1993a). Before their breakdown the populations of ribosomes, dictyosomes and RER cisternae diminish in the tapetal cells, and lipid bodies and plastids become conspicuous features of the cytoplasm. Eventually the tapetal cell membrane degenerates and the contents enter the locule and are deposited on the pollen surface as pollenkitt/tryphine. In many taxa characterized by secretory tapetum, sporopollenin granules, termed ‘Ubisch bodies’ or ‘orbicules’ are deposited (Fig. 1.3) on the inner tangential surface of the tapetal cells (Pacini and Franchi 1992,1993). Orbicules origi­ nate in the cytoplasm of the tapetal cells as lipoidal pro-orbicular bodies covered with a mem­ brane. Pro-orbicular bodies accumulate below the membrane and eventually extrude to the cell surface (facing the locule) where they acquire a sporopollenin coating. Orbicules are absent in

taxa characterized by plasmodiai tapetum and also in a few taxa with a secretory tapetum (Murgia et al. 1991, Pandolfi et al. 1993).

Plasmodiai Tapetum The inner tangential and radial walls of the tapetal cells break down and the tapetal proto­ plasts intrude into the thecal cavity amid the microspores. In advanced monocots the intru­ sion generally occurs earlier, while in primitive monocots and dicots the intrusion occurs at the microspore stage (Pacini et al. 1985). The iden­ tity of individual tapetal protoplasts may be main­ tained or the protoplasts may fuse in the locules to give a periplasmodium (Fig. 1.7A, B) (Tiwari and Gunning 1986a-c, Pacini and Keijzer 1989). Ultrastructural studies of the plasmodiai tape­ tum indicate that the protoplasts/periplasmodium are/is not a degeneration product but an orga­ nized and functional unit with normal organelle distribution (Pacini and Juniper 1983, Tiwari and Gunning 1986a-c, Fernando and Cass 1994, Galati 1996). An increase in number of mitochon­ dria and elaboration of ER membrane, reported in tapetal protoplasts/periplasmodium, is indica­ tive of their/its high metabolic activity (Fernando and Cass 1994, Galati 1996). The plasmodiai tapetum comes into close contact with the mi­ crospores/young pollen grains, facilitating trans­ port of substances from the tapetum to the mi­ crospores. Towards the end of pollen develop­ ment the tapetal protoplasts/periplasmodium degenerate(s) and adhere(s) to the pollen sur­ face as pollenkitt. Sporopollenin bodies (orbicules) are gener­ ally absent in the plasmodiai tapetum. However, sporopollenin bodies have been reported in a few species with plasmodiai tapetum such as Tradescantia (Tiwari and Gunning 1986a, b) and Butomus (Fernando and Cass 1994). In Butomus vesicles produced by extensive arrays of RER in tapetal cells/protoplasts have been implicated in the formation of sporopollenin-like bodies and sporopollenin precursors.

Pollen Development 21

Tapetal Membrane Development of tapetum is associated with for­ mation of an acetolysis resistant membrane, termed the tapetal membrane in many species of angiosperms as well as gymnosperms (Heslop-Harrison 1972). This appears to be a general feature of seed plants. In species char­ acterized by a secretory tapetum, the tapetal membrane is generally formed on the inner sur­ face of the tapetal cells (towards the locule) while in species characterized by a plasmodiai tape­ tum, the membrane is formed on the outer sur­ face of the tapetum (towards the endothecium) (see also Parkinson and Pacini 1995).The mem­ brane is largely made up of sporopollenin; in­ soluble polysaccharides such as cellulose, cal­ lose and pectin also seem to be present in small amounts (Gupta and Nanda 1972). The tapetal membrane forms a culture sac around the pollen at a later phase of its devel­ opment. The function of the tapetal membrane is not clearly understood (Shivanna and Johri 1985). As it is largely made up of sporopollenin, it may restrict the free passage of materials into and out of the pollen mass.

Role of Tapetum Much direct and indirect evidence clearly shows that the tapetum plays a crucial role in pollen development. Pollen sterility (nuclear/cytoplas­ mic/environmental) is invariably associated with tapetal abnormality (see Chapter 5). It has also been possible to induce pollen sterility by tar­ geting the tapetum through recombinant DNA technology (see Chapter 5). The role of the tape­ tum in pollen development seems to be mani­ fested largely after completion of meiosis. The major functions attributed to the tapetum are described below.

Supply of nutrients to developing pollen The tapetum has traditionally been considered to be the nurse tissue for the developing pollen. As the tapetum encloses the sporogenous tis­ sue all around, any nutrients entering the

sporogenous cells have to pass through the tapetum. In the secretory tapetum, metabolites in the form of soluble carbohydrates, amino acids and peptides are released into the locular fluid through exocytosis or secretion, from where they are taken up by the developing pollen (Pacini 1994). In the plasmodiai tapetum, the plasma membrane of the tapetal protoplasts are closely adpressed to the developing pollen mak­ ing the passage of nutrients more efficient than in the secretory tapetum. The build-up of nutri­ tional reserves in the pollen is generally associ­ ated with breakdown of the tapetum. In Sorghum and many other members of Poaceae the pore end of the pollen grains lies adjacent to the tapetal cells. Starch build-up in the pollen is ini­ tiated at the germ pore end of the pollen, indi­ cating absorption of tapetal products through the pore (Christensen and Horner 1974). In many species the intine in the poral region pro­ duces wall ingrowths similar to transfer cells (Charzynska et al. 1990). The wall ingrowths have been suggested to facilitate nutrient and water uptake initially from the tapetum and sub­ sequently from the anther locule. In Acacia also (McCoy and Knox 1988) membrane labyrinths present in the apertural intine have been impli­ cated in the transport of metabolites.

Breakdown of callose wall around microspore tetrads The enzyme callase, required for breakdown of the callose wall around the microspore tetrads, is released by the tapetum. The microspores themselves are incapable of callase synthesis; culture of isolated microspore tetrads does not result in breakdown of callose. The tapetal plas­ modium, however, is able to break down the callose of microspore tetrads as well as sieve tubes (Mepham and Lane 1969). Biochemical analysis of callase in the developing anther has shown that the dissociation of tetrads is corre­ lated with a sharp increase in callase activity; most of the activity is localized in the wall of the anther and not in the meiocytes (Stieglitz and

22 Pollen Biology and Biotechnology

Stern 1973, Scott et al. 1991). In Pterostylis a proteinaceous layer is deposited in the inner tan­ gential wall of tapetal cells at the tetrad stage (Fitzgerald et al. 1993a). However, the wall be­ comes uniformly thin following microspore re­ lease. It is suggested that the material depos­ ited in the tapetal wall is callase and is released before callose breakdown (Fitzgerald et al. 1993b). Tapetal activation of callase at the right stage is very important for normal development of pollen. Mistiming of callase activation results in pollen sterility (see Chapter 5).

Supply of sporopollenin precursors to pollen exine Tapetum is considered to provide sporopollenin precursors for exine formation. Although the blueprint of exine is laid down before the mi­ crospores are released, in most taxa the bulk of the exine is deposited after release of the mi­ crospores (Echlin 1971, Heslop-Harrison 1971, Owens and Dickinson 1983). The orbicules pro­ duced in the secretory tapetum seem to be the end products of metabolism; so far no enzyme capable of degrading sporopollenin has been identified. Orbicules, therefore, do not seem to take part in the formation of exine. The precur­ sors secreted by the secretory as well as plasmodial tapetum are likely to be involved in exine formation (Heslop-Harrison 1971, Horner and Pearson 1978, Owens and Dickinson 1983, Fernando and Cass 1994). Available evidence suggests that the development of ektexine pro­ ceeds centrifugally through the deposition of tapetally derived sporopollenin; the endexine develops centripetally by the deposition of sporopollenin derived from pollen protoplast, deposited on membrane-like strands. In Butomus (Fernando and Cass 1994) the tapetal cells, before formation of the periplasmodium, extrude particles into the locule that condense to form electron dense aggregates. These ag­ gregates are similar to the primexine and are incorporated into the developing ektexine. Sporopollenin produced from the tapetum is laid

down on the developing ektexine even after the tapetum becomes plasmodiai and completely encloses the microspores.

Supply of pollen coat substances and exine proteins The tapetal origin of pollen coat substances generally known as pollenkitt or tryphine is well established (see Chapter 2). The proteins are located in exine cavities (tectate grain) or in the surface depressions (non-tectate grain). During meiosis, proteins and lipids accumulate in tapetal cells, the former in single membrane-bound vesicles apparently derived from RER and the latter in plastids. Following the breakdown of tapetal cells, these proteins and lipids are re­ leased into the thecal cavity and get deposited in the exine. The lipoidal components generally remain on the surface of the pollen (HeslopHarrison et al. 1973,1974).The pollenkitt is pre­ dominantly made up of lipids, flavonoids and carotenoids, and other degenerated products of the tapetum (Wiermann and Vieth 1983; see Chapter 2), and plays an important role in pol­ len dispersal and pollen function. Tryphine is often distinguished from pollenkitt; the former is a complex mixture of hydrophilic substances derived from the breakdown of tapetal cells, while the latter is principally the hydrophobic lipids containing species-specific carotenoids (Echlin 1971, Dickinson 1973). In Pterostylis pollenkitt acts as a glue to bind pollen grains into a pollinium; in this species there are no ex­ ine bridges between pollen grains (see Chapter 2 for details).

GENE EXPRESSION DURING ANTHER DEVELOPMENT Gene expression during anther and pollen de­ velopment has been an active area of research in many laboratories and considerable informa­ tion has accumulated during the last 15 years. An excellent account of the available data is given by Mascarenhas (1990,1992), Goldberg et al. (1993), McCormick (1993), Hamilton and

Pollen Development 23

Mascarenhas (1997) and Raghavan (1999).The following is a brief summary: Over 24,000 different RNAs have been re­ ported in mature pollen. About 65% of these genes are also expressed in the sporophytic tis­ sues and the remaining (ca. 10,000) seem to be expressed exclusively or predominantly in the anther and pollen. The expression of these genes is regulated both temporally and spatially. A number of anther-specific genes and/or their cDNAs have been isolated. Several genes (such as Ta 13, Ta 26, Ta 29) are expressed only in the tapetum, while some are expressed in the con­ nective tissue, endothecium and middle layers {[Ta20,Ta55). Anther-specific mRNAs have been shown to encode a range of proteins such as lipid transfer proteins, protease inhibitors, thiol endopeptidase, glycine-rich and proline-rich polypeptidases, pectate lyases, and chalcone synthase. Comprehensive studies have been carried out on the temporal expression of genes during pollen development. The genes expressed in pollen have been grouped as ‘early’ and ‘late’ genes. The mRNAs of early genes are first de­ tectable soon after meiosis, reaching a maxi­ mum at late pollen interphase, and decreasing thereafter. Expression of late genes starts at or after microspore mitosis and reaches maximum at the mature pollen stage. Late genes form the bulk of the pollen-specific genes identified to date. Early genes are presumed to be associ­ ated with pollen development, while late genes are likely to play a role in pollen maturation, ger­ mination and pollen tube growth (Mascarenhas 1990). Numerous cDNAs have been character­ ized, in particular of ‘late’ genes. Some of these are BP 10 (Brassica napus), Bcp 1 (B. campestris), NTP 303 (tobacco), LAT 51 (to­ mato), Zm 13 (maize), Org cl (rice) and Bet v 1 (white birch). LAT 51 sequences seem to be highly conserved as similar sequences are found in diverse plant species. Most of the putative proteins of different genes exhibit sequence homologies to known proteins such as wall-de­ grading enzymes, cytoskeleton proteins and

allergens (Hamilton and Mascarenhas 1997). One of the genes, Bcp 1, expressed in the tape­ tum as well as microspores (Theerakulpisut et al. 1991), is essential for pollen fe rtility in Arabidopsis (Muschietti et al. 1994, Xu et al. 1995a). Pex 1 (pollen extensin-like) gene in maize, expressed exclusively in pollen, has an extensin-like domain (Rubinstein et al. 1995). A pollen-specific gene in rice (PS 1) shows sig­ nificant levels of homology to ‘late’ expressed maize Zm 13 and tomato LAT 52 (Zou et al. 1994).

TRANSMISSION OF PATHOGENS THROUGH POLLEN Only limited studies have been carried out to determine whether pathogens can pass through pollen grains to the progeny. There are some reports of transmission of viruses through pol­ len (Mandahar 1981, Mink 1993). Electron microscopic studies of pollen grains from infected plants have shown the presence of viruses (Carroll 1974, Hamilton et al. 1977). Transmission of barley stripe mosaic virus via pollen has been dem onstrated through ultrastructural studies of pistils of healthy plants pollinated with virus-infected pollen (Brlansky et al. 1986). Virus particles were detected not only in the pollen tubes but also in the zygote and the resultant embryo and endosperm. Although some early studies (Ark 1944) reported pollen as a source of some bacterial infection, there are no conclusive reports on the transmission of bacterial diseases through pollen.

POLLEN MATURATION AND ANTHER DEHISCENCE Towards the end of its development, pollen builds up reserve materials (starch/lipids).The plastids differentiate into amyloplasts (Pacini et al. 1992, Franchi et al. 1996). In many species mature pollen grains contain starch grains (starchy pol­ len), while in others starch is hydrolyzed before pollen grains are shed (starchless pollen). On

24 Pollen Biology and Biotechnology

the basis of analyses of pollen of 124 families, Baker and Baker (1979) found that primitive fami­ lies tend to have starchy pollen and advanced families starchless pollen. Also, entomophilous species pollinated by Hymenoptera and Diptera generally produce starchless pollen and those pollinated by Lepidoptera as well as birds pro­ duce starchy pollen. Starchy pollen grains tend to be larger than starchless pollen (Endress 1994). Irrespective of the nature of reserve material present in mature pollen, plastids invariably de­ velop starch either during microspore develop­ ment (after microspore mitosis) and/or during maturation (Franchi et al. 1996, Speranza et al. 1997). In some species the starch present in amyloplasts is the only carbohydrate source in mature pollen and the cytoplasm has no PASpositive material (eg. Lolium, Cucurbita). In others amyloplast starch is partially hydrolyzed and the cytoplasm contains low molecular weight insoluble PAS-positive material (eg. Dactylis, Cucumis). In several species starch is completely hydrolyzed during maturation and mature pol­ len is starchless; but the cytoplasm contains PAS-positive material and simple sugars such as glucose and fructose (e.g. Borago, Lycopersicon). Recent survey of pollen grains of 901 spe­ cies belonging to 104 dicot and 15 monocot fami­ lies (Franchi et al. 1996, Pacini 1997) showed variations in physicochemical properties of starch in the intensity of colour development fol­ lowing l-KI staining and the presence or absence of birefringence under polarized light. The re­ sults also indicated that starchless pollen grains withstand desiccation better than starchy pollen. Desiccation tolerance has been attrib­ uted to low molecular weight carbohydrates in the cytoplasm of starchless grains (see Chapter 3). Anther dehiscence is the result of precise co-ordination between differentiation and/or de­ generation of cell layers of the anther and dehy­ dration. Although extensive studies have been carried out on the details of pollen development, only limited information is available on the struc­

tural and physiological details associated with anther dehiscence (Keijzer 1987a, b, Bonner and Dickinson 1989, Goldberg et al. 1993). In one of several comprehensive studies on anther dehis­ cence, Bonner and Dickinson (1989) have pre­ sented structural and cytochemical aspects of anther dehiscence in Lycopersicon esculentum. The earliest event associated with anther de­ hiscence is observed in the cells of the intersporangiai septum (ISS) (Bonner and Dickinson 1989) or circular cell cluster (Koltunow et al. 1990). In members of Solanaceae, a large number of calcium oxalate crystals accumulate in the cells of the ISS followed by their degen­ eration. However, in many other species such as Gasteria, degeneration of the cells of the ISS is not accompanied by crystal formation. ISS de­ generation is associated with activation of many hydrolytic enzymes such as cellulases, pectinases and acid phosphatases. Degenera­ tion of cells of ISS results in rupture of the sep­ tum leading to the fusion of the two locules. De­ generation of ISS is also associated with differ­ entiation of the stomium in the epidermal layer along the line of dehiscence. Unlike the neighbouring epidermal cells which show en­ largement with thickened cuticle, the stomium cells remain small and are covered with thin cuticle, thus creating a point of weakness in the anther wall. The protoplasts of cells of the stomium and epidermis degenerate. Simulta­ neous with degeneration of the epidermal layer, cells of the endothecium enlarge and develop characteristic thickenings. Eventually the cells of the anther wall desiccate and collapse, re­ sulting in a narrow split along the stomium. Using the strategy of ablation of specific cell types, Beals and Goldberg (1997) showed that anther dehiscence depends on the presence of a functional stomium. Ablation of cells in the stomium region leads to failure of anther dehis­ cence. Desiccation of anther tissue seems to re­ sult from a combination of active resorption of water through anther filament and cuticular evaporation (Heslop-Harrison et al. 1987, Pacini 1994). Removal of anthers from their filaments

Pollen Development 25

delayed anther dehiscence when compared to those still attached to the flowers (Schmid and Alpert 1977). In many members of Poaceae, an­ ther filaments show remarkable elongation just before anther dehiscence and bring the anthers above the level of other floral organs. During this elongation water is retracted from anthers and moved to the filaments (Keijzer et al. 1987, Keijzer 1999); this results in synchronous dehy­ dration of anthers and supply of water to the

vacuoles of the filament cells facilitating their elongation. Dehiscence requires activation of many genes, particularly those that encode hydrolytic enzymes. One of the genes TA 56 (thiol pepti­ dase) in tobacco seems to be involved in the dehiscence process (see Goldberg et al. 1993). TA 56 mRNA accumulates first in the ISS prior to its destruction, subsequently in the stomium and finally in the connective.

2 Pollen Morphology and Aeropalynology POLLEN MORPHOLOGY Pollen morphology is of great significance in taxonomy, phylogeny, palaeobotany, aeropa­ lynology and pollen allergy. Analysis of fossil pollen is the most important approach to reconstruction of past flora, vegetation and environment (Faegri and Iversen 1989). Pollen morphology is also important in understanding the functional aspects of pollen such as pollination biology and pollen-pistil interaction. Pollen identification, the basis of palynology, is based exclusively on pollen morphology. This chapter provides a brief description of technical terms used in pollen morphology. Several books (Erdtman 1966, 1969, Nair 1970, Moore and Webb 1978, Nilsson and Muller 1978, Thanikaimoni 1978, Faegri and Iversen 1989) are available for a comprehensive coverage of pollen morphology and pollen analysis.

Applications Petroleum exploration One of the most important applications of pa­ lynology is in the field of petroleum exploration. Until the first quarter of the 20th century, oil ex­ ploration was largely confined to the recovery of oil and gas in shallow deposits. Potential sources were identified on the basis of rock strata. However, these approaches proved in­

sufficient for exploring oil in deeper wells. Analy­ sis of fossil pollen present in rock samples col­ lected from various depths in oil wells proved very useful in predicting oil and gas zones. Pres­ ently almost all oil exploration companies have well-established palynology laboratories to as­ sist oil exploration work (Faegri and Iversen 1989, Bryant 1990).

Archaeology Pollen analysis has important applications in understanding past climatic changes, the origin and spread of agriculture and prehistoric cul­ tures (Bryant 1990). By comparing fossil pollen records of archaeological sites with well-dated pollen records, archaeologists are often able to date specific archaeological events. Analysis of pollen in archaeological human burials and samples of coprolites (preserved faeces) has given definitive information about ancient diets.

Criminology Unlike many other macroevidences, which can be easily removed by criminals, pollen grains cannot be removed easily. Analysis of such pol­ len can be effectively used in solving crimes (Bryant 1990). For example, in one of the wellpublished murder trials in Austria in 1959 evi­ dence from pollen grains was responsible for identifying the murderer. A man disappeared near Vienna while on a journey down the Danube

Pollen Morphology and Aeropalynology 27

River, but his body was not found. The man who had a motive for killing was arrested and charged with murder. In the absence of confession or the body, prosecution could not prove the murder. Analysis of pollen in the mud found on the defendant’s shoes revealed the pollen of spruce, willow and alder together with fossil hickory pol­ len (from exposed Miocene-age deposits). Only the soil from one small area 20 km north of Vienna along the Danube River valley contained this pollen mixture. When confronted with this information the shocked defendant confessed his crime and showed the authorities the site where the body was buried which turned out to be the region pinpointed on the basis of pollen analysis (Bryant 1990). Although pollen can pro­ vide definitive clues in solving crime, it is not yet being used widely, largely because of lack of sufficient basic pollen data of different regions and of experts to identify pollen grains.

Testing purity of honey Honey-bees collect nectar and pollen from plants available around the hive and use them to make honey. Thus analysis of pollen present in the honey indicates not only the geographic region of honey production, but also the plant species which have been used.

Three Domains of Pollen Wall The pollen wall is probably the most complex wall system found in higher plants. It includes three main domains—intine, exine and pollen coat—which differ in morphology, chemistry and function.

Intine The intine envelopes the pollen protoplast and is comparable to the primary wall of other plant cells. It is primarily composed of cellulose, hemicellulose and pectic polymers. In some taxa such as grasses, a middle layer between the exine and intine rich in pectic polysaccharides, termed the Z-layer, is distinguishable. The Z-layer is thickened at the germ pore region and is termed

‘Zw ischenkorper’ (Rowley 1964, HeslopHarrison 1979c). The intine invariably contains protein in the form of radially elongated tubules generally concentrated in the germ pore region (see page 33).

Exine The exine layer is highly sculptured and orna­ mented (Fig. 2.1). It is composed of sporopolle­ nin, a highly resistant organic biopolymer. Viscin threads that bind together pollen grains and tet­ rads in some families such as Onagraceae and Ericaceae are also made up of sporopollenin (Hesse 1981, 1984). Because of the difficulties in purification of sporopollenin and its insolubil­ ity in most solvents, sporopollenin is not readily amenable to chemical analysis. Since the 1960s sporopollenin has been considered a biopoly­ mer derived from carotenoids and carotenoid esters (Brooks and Shaw 1971). However, re­ cent comprehensive studies involving tracer and degradation experiments and spectrometric analysis have indicated sporopollenin to be a mixed polymer with a large amount of aliphatics containing additional compounds such as phenols. Pollen identification is largely based on the structure and surface sculpturing of the exine. Two sets of terminologies on exine stratification are in vogue (Faegri and Iverson 1989, Erdtman 1969) (Fig. 2.2). On the basis of chemical resis­ tance, autofluorescence and staining capacity, two layers of exine, ektexine and endexine are distinguished (Kress and Stone 1982). The ek­ texine dissolves with 2- and 3 -ethanolamine and stains deeply with alcoholic fuchsin and auramine O; the endexine resists hydrolysis with ethanolamine and stains weakly or not at all with alcoholic fuchsin and auramine O. The endex­ ine forms a homogeneous layer and is continu­ ous in the non-apertural region in most pollen. However, in some taxa, such as grasses, there are fine passages across the endexine (Rowley 1973, Dahl and Rowley 1974).

28 Pollen Biology and Biotechnology

Fig. 2.1 Scanning electron micrographs of some pollen grains to show variations in exine ornamentation.

Pollen Morphology and Aeropalynology 29

B tectum ine - j ^ sexine — ektexine —

colum ella exine nexine

— exine ----------- fo o t layer

j - nexine 1 -------- -----------------------------------------------------I—nexine 2 ----- ------------------------------------------------------------in tin e ---~ ■;•;

----- intine

Fig. 2.2 Pollen wall architecture and two sets of terminologies (A, B) used to describe exine layers.

Pollen coat substances Pollen grains, especially of entomophilous plants, are coated with an oily, sticky and often coloured material commonly termed ‘pollenkitt’ or ‘tryphine’. In other species, it is less conspicu­ ous. The pollen coat substances contain a range of chemicals; lipids, carbohydrates, proteins, gly­ coproteins, carotenoids, and flavonoids; these and other phenolics are the major components of the pollen coat (Wiermann and Gubatz 1992). Pollen coat substances are derived from the tapetum. The carotenoids are synthesized in the tape­ tum and accumulate in osmiophilic globules (Heslop-Harrison and Dickinson 1969). After the breakdown of the tapetum, the carotenoids are released into the locule and eventually get de­ posited on the exine surface and/or cavities. Flavonoids are important phenolic compounds of the pollen coat substances. Kaempferol, quercetin and isorhamnetin are the principal fla­ vonoids. Flavonoid biosynthesis occurs largely in the tapetum (Wiermann and Gubatz 1992). L-phenylammonia-lyase (PAL) and chalcone synthase (CHS) are the key enzymes in fla­ vonoid biosynthesis; both these enzymes are predominantly distributed in the tapetum. It is suggested that the enzymes involved in flavonoid biosynthesis are released into the locule and are transported to the pollen where flavonoid syn­ thesis takes place. In Brasslca napus oleosin-

like proteins, major pollen coat components, are synthesized in the tapetum (Piffanelli et al. 1997, Murphy and Ross 1998).

Pollen Wall Morphogenesis The wall of the pollen grain is rather uniform in architecture (see Fig. 2.2 for details). Pollen wall morphogenesis occurs in two stages. The first stage is completed when the microspores are still enclosed by a callose wall and the second after the release of microspores (HeslopHarrison 1963). Details of pollen wall morpho­ genesis are diagrammatically represented in Figure 2.3. This scheme of exine deposition, generally referred to as the primexine scheme (Heslop-Harrison 1963), has been extensively adopted in studies on pollen wall development. The blueprint of exine, termed primexine, is laid down between the plasma membrane and the callose wall. The primexine has a matrix, pre­ sumably made up of cellulose, and radially di­ rected rods, the probaculae. Probaculae become connected at their bases to form the foot layer. The probaculae may remain free above or form a roof over the primexine matrix. The foot layer represents the future nexine 1, and the probaculae and the roof layer (tectum) the fu­ ture sexine (Erdtman 1966,1969).The primexine and foot layer are acetolysis resistant. Since the electron opacity of the probaculae and foot layer differs from that of mature exine, the com­ ponent of these layers has been termed

30 Pollen Biology and Biotechnology

‘protosporopollenin’. In Dendrobiumfreeze frac­ ture and freeze substitution studies have shown tubular SERs terminating at and fusing with the plasma membrane (Fitzgerald et al. 1994). It has

tion. It is suggested that ER physically prevents the movement of membranous structures coated with primexine material to the cell surface (Dickinson and Potter 1976, Sheldon and Dickinson 1983). In a few species, however, pre­ sumptive pore regions are not associated with ER localization (Shoup et al. 1980). In a number of species the position of the probaculae and other features of the mature exine can also be correlated with specific fea­ tures in the cytoplasm of young microspores. For example, the location of probaculae in Silene (Heslop-Harrison 1963) and Berberis (Gabara 1974) is marked by the presence of short mem­ brane profiles below the plasma membrane, and of those of Lilium (Vazart 1970) by accumula­ tion of ribosomes. In Cosmos (Dickinson and Potter 1976), large banks of ER and microtu­ bules in the cytoplasm are associated with the development of spines of the exine. Microspores are released by the dissolution of callose wall soon after the differentiation of the primexine. Recent studies on exine development in Caesalpinia and Lilium (Takahashi 1989,1993, 1995), and a few other taxa (Takahashi and Skvarla 1991, Fitzgerald and Knox 1995) do not support the primexine scheme of exine deposi­ tion. These studies have shown that the exine pattern is initiated by invaginations of the mi­ crospore plasma membrane at an early tetrad stage (Fig. 2.4). In Lilium (Takahashi 1995), the invaginated plasma membrane takes the form of a reticulate pattern that corresponds to the pattern of mature exine.The invaginated regions correspond to the regions of future lumina and

been suggested that the enzymes for polymer­

protubérants (raised areas) correspond to the

ization of sporopollenin precursors are trans­ ported via SER to the plasma membrane. Ac­ cording to the primexine scheme, the structural features of exine are determined by the distri­ bution pattern of probaculae. In many taxa the presumptive germinal ap­ ertures (pores/colpae) are demarcated during the formation of primexine by the presence of ER oriented parallel to the plasma membrane. These regions remain free of primexine deposi­

muri of mature exine. At the regions of raised areas of the plasma membrane, fibrous threads of 10-20 nm diameter aggregate together with granules of 10 nm diameter. Gradually the ag­ gregated fibrous threads and granules develop into 0.5-0.7 pm wide smooth protectum. Probaculae form subsequently below the protectum (between the protectum and the plasma membrane). At the late tetrad stage the protectum and probaculae become distinct

callose wall plasma membrane

primexine probacula

exine intine


E Fig. 2.3. Diagram of pollen wall morphogenesis based on primexine model.

Pollen Morphology and Aeropalynology 31

-callose wall -plasma membrane

'»-----plasma membrane


'S---- protectum

-------- degrading callose — probacula ■primexine


Fig. 2.4 Diagram of pollen wall morphogenesis based on plasma membrane undulation model.

below the callose wall and the plasma mem­ brane becomes smooth. The fibrous primexine matrix becomes visible in the spaces between probaculae. The callose wall dissolves at this stage, releasing the microspores. At the time of release the surface of the protectum is smooth. Further differentiation of exine continues after microspore release. Thus the reticulate pattern of the invaginated plasma membrane forms the blueprint of mature exine. The protectum is the first exine layer to be deposited on the reticulate patterned plasma membrane. According to this scheme, plasma membrane plays a very impor­ tant part in pollen wall morphogenesis. However, the mechanism by which plasma membrane de­

velops patterned invaginations is not known. It is possible that cytoskeleton elements play a role in this differentiation. Studies of Fitzgerald and Knox (1995) on pollen wall initiation in Brassica microspores showed basically the same pattern of plasma membrane undulations as described above. However, before the undulations became visible continuous primexine deposition could be ob­ served between the microspore plasma mem­ brane and callose wall. Deposition of flakes of condensed material was observed in the depres­ sions of the plasma membrane before deposi­ tion of the probaculae; this material represents interbaculate sites, referred to as spacers. This type of exine morphogenesis (based on undu­ lations of plasma membrane) seems to be a fea­ ture specific only to reticulate exine. In species such as Acacia (Fitzgerald et al. 1993a) and Dendrobium (Fitzgerald et al. 1994) which do not have reticulate exine, plasma membrane undulations are not associated with exine depo­ sition.

Control of exine pattern Exine ornamentation is one of the unique fea­ tures of pollen. Although there is great variation in exine ornamentation between species it is characteristic of a given species/genus. Initial observations that the blueprint of the exine is laid down when the microspores are still en­ closed by the callose wall led to the suggestion that the exine pattern is under the control of the gametophytic genome (Godwin 1968). However, subsequent studies have clearly shown that the exine pattern is under the control of the sporophytic genome (Heslop-Harrison 1971, 1972). Some of the evidence is presented below. In an interspecific hybrid of Linum (Rogers and Harris 1969), many chromosomes are left in the cytoplasm during meiotic division because of irregular chromosomal pairing. During cleav­ age of the microspore mother cell, small spores form around the scattered chromosomes. All these spores develop typical exine although they have an incomplete chromosome complement

32 Pollen Biology and Biotechnology

(Fig. 2.5). If the exine pattern is to be under the control of the microspore genome, typical exine formation in all such spores cannot be expected, as it is not possible for all these spores to re­ ceive the genome complement necessary for control of the exine pattern. Development of typi­ cal exine in microspores with incomplete chro­ mosome complement has also been reported in triploid clones of Tradescantia (Mepham and Lane 1970) and male sterile mutants of Impa­ tiens (Tara and Namboodiri 1974). It has been a common observation that there is no segregation of exine pattern in pollen grains of hybrids of parents showing differences in exine features. For example, in the hybrid

Lycopersicon esculentum x Solanum pennellii (Quiros 1975), the genes of S. pennellii are domi­ nant for pollen size and exine density, and those of L. esculentum for pollen shape and exine spine structure. There is no segregation of ex­ ine pattern in the pollen grains of the hybrid. Although all the above evidence confirms sporophytic control of the exine pattern, there is no clear information on morphogenetical details of such control. Apparently the information is transcribed before meiosis and is carried in the cytoplasm of the meiocytes in the form of longlived mRNAs or protein moiety (Heslop-Harrison 1972). It is possible that long-lived mRNAs and/ or proteins are retained in the cytoplasm en­

Fig. 2.5 Anomalous development of pollen grains in Linum hybrid. A. Metaphase I with several scattered laggards. B-D. Microspore formation resulting in a tetrad (C) or a dyad (D) accompanied by several smaller spores. E-G. Pollen grains of various sizes with fully developed exine (after Rogers and Harris 1969).

Pollen Morphology and Aeropalynology 33

closed by double or multimembrane units (see Chapter 1), which are not affected by the cyto­ plasmic reorganization that takes place in other parts of the cytoplasm (Dickinson and HeslopHarrison 1977).

Pollen wall proteins It is now well established that both layers of the pollen wall, the exine and intine, contain a con­ siderable amount of proteins (Figs 2.6, 2.7) (Heslop-Harrison 1975, Heslop-Harrison et al. 1975, Knox et al. 1975, Shivanna and Johri 1985). Exine proteins are located in the interporal regions in exine cavities (in tectate pollen) or surface depressions (in pilate pollen). Exine pro­ teins originate from the surrounding tapetum and thus are sporophytic in origin (Fig. 2 .8). The in­ tine proteins are generally present in the form of radially oriented tubules and are concentrated in the germ-pore region. Intine proteins originate in the pollen cytoplasm (Fig. 2.8) and are there­ fore gametophytic in origin. Soon after the initia­ tion of intine development, the plasma mem­ brane of pollen produces radially oriented tu­ bules into the developing intine. Eventually, these tubules with their protein inclusions are cut off from the plasma membrane and become incor­ porated in the intine. Pollen wall proteins include many enzymes, particularly hydrolytic enzymes such as esbacula---------------columella -------tectum exine proteins microspore ■ — exine — ■ ■■ ------- intine--------------- intine proteins -

plasma membrane .

- surface lipids —

Fig. 2.6 Location of intine and exine proteins, and pollen coat lipids. A. Tectate exine. B. Pilate exine (modified from Shivanna 1977).

terases, amylases and ribonucleases (Knox and Heslop-Harrison 1969,1970b). Esterases occur predominantly in the exine and acid phos­ phatases in the intine. These enzymes can there­ fore be used as marker enzymes for exine and intine proteins. Proteins responsible for pollen allergy are also present in the pollen wall. Pol­ len wall proteins play an important role in pollen function (see Chapters 8 , 9). In many species belonging to Cannaceae, Zingiberaceae, and marine angiosperms in which exine is absent or highly reduced, tapetally derived proteins are absent. The intine is gener­ ally thick and often differentiated into many lay­ ers and the wall proteins are exclusively con­ fined to the intine (Heslop-Harrison 1975, Knox 1984a, Kress et al. 1978, Kress and Stone 1982, Duckeretal. 1978). Apart from the deposition of tapetally derived proteins in the interporal region of pollen, in many species such as Olea (Pacini and Juniper 1979, Pacini et al. 1981) and Pterostylis (Pandolfi et al. 1993) tapetal proteins are also deposited in the poral region (where exine is absent) out­ side the intine (Shivanna and Johri 1985). Un­ like interporal exine proteins which are depos­ ited following tapetal breakdown, poral proteins are deposited earlier, when the tapetal cells are still intact.

Pollen Analysis Pollen size and shape The size of the pollen is highly variable and ranges from 5 to 200 pm. Smallest grains are recorded in members of Boraginaceae and larger pollen grains are produced in mem­ bers of Cucurbitaceae, Malvaceae and Nyctaginaceae. The pollen of Cymbopetalum odoratissimum (Annonaceae) measuring up to 350 pm is perhaps the largest reported (Walker 1971). In a majority of species pollen size var­ ies between 15 and 50 pm. In marine an­ giosperms such as Amphibolis and Zostera, pollen grains are filiform and long measuring up

34 Pollen Biology and Biotechnology

Fig. 2.7 Transmission electron micrographs of part of a pollen wall to show incorporation of pollen wall proteins. A, B. Iberis. A. Microspore before insertion of tapetal material in exine depressions. The tapetum (t) is still intact and the intine (i) is not fully developed. B. After insertion of the tapetum-derived components in the cavities of the exine. C, D. Crocus characterized by thick intine following localization of acid phosphatase activity. The enzyme is localized in radially oriented tubular inclusions in the intine. D. Part of the intine magnified to show details (e— exine, i—intine, c—cytoplasm, I—lipids, t—tapetum, tp— tapetal proteins) (after Heslop-Harrison 1975).

Pollen Morphology and Aeropalynology 35

plasma membrane lipids proteins tapetum

-microspore - tapétum cytoplasmic^ extensions -

intine proteins—

exine proteins-


Fig. 2.8 Origin of intine and exine proteins. A. Transection of a young anther with microspores and intact tapetum. Only part of the microspore wall and tapetum are shown in B-D to indicate incorporation of proteins during pollen development. Exine proteins are derived from the tapetum and intine proteins from the pollen cytoplasm (after Shivanna 1977).

to 5 mm (Ducker et al. 1978). In Crossandra, a terrestrial species, pollen grains are long (>500 pm) (Brummitt et al. 1980). The size of pollen grains varies to some ex­ tent, depending on the method of preparation and its hydration level.Therefore, pollen size can hardly serve as a diagnostic character (Faegri

and Iversen 1989). However, with other morpho­ logical characters, size of the pollen may help in demarcating the taxa. The shape of the pollen is also variable. Des­ iccated grains are frequently subprismatic be­ cause of the contraction of the exine, while hy­ drated grains tend to be ellipsoidal or spherical.

36 Pollen Biology and Biotechnology

Also the shape of pollen grains varies depend­ ing on whether it is observed in polar view or equatorial view. Pollen grains of a large number of species in polar view are circular or elliptic. However, triangular, quadrangular and rectan­ gular pollen are also present. In angular pollen, the angles may be acute or obtuse (Moore and Webb 1978, Faegri and Iversen 1989).

Compound pollen Associations of two or more pollen grains as dis­ persal units are referred to as compound or com­ posite pollen grains (Fig. 2.9) (Walkerand Doyle 1975, Knox and McConchie 1986, Pacini 1997).

The number of pollen grains in a compound unit may be 2 (dyad), 4 (tetrad) or multiples of four (polyad). In several species, a large number of pollen grains associate to form massulae (some members of Orchidaceae). In a number of Asclepiadaceae and Orchidaceae all the mi­ crospores in a sporangium remain together to form a single mass called the pollinium. Com­ pound pollen grains have been reported in more than 56 angiosperm families (Walker and Doyle 1975). Compound pollen may be formed by fu­ sion of their exine layers (calymmate) or by con­ necting wall bridges (acalymmate). Wall bridges may be confined only to exine layers (sexine and

CD Dyad





Fig. 2.9 Diagram of compound pollen grains (modified and redrawn from Pacini 1997).

Pollen Morphology and Aeropalynology 37

nexine) or include intine also. Some species show cytoplasmic bridges between individual units of the compound pollen.

Polarity The morphological terms used to describe pol­ len are far too many and often confusing (Kremp 1965, Nilsson and Muller 1978, Thanikaimoni 1978).The following paragraphs provide only the basic terms used to describe pollen. Pollen mor­ phology is largely described in relation to the polarity of the pollen grain. The polarity of the pollen is referenced to the arrangement of mi­ crospores in the tetrad (Fig. 2.10).The pole clos­ est to the center of the tetrad is the proximal pole and that farthest from the center is the dis­ tal pole. A hypothetical axis connecting the two poles is the polar axis. The axis at a right angle to the polar axis is the equatorial axis. It demar­ cates the two halves of the pollen; if the two halves are similar, the pollen grain is described as isopolar, and if different, as heteropolar. Even after the release of microspores from the

tetrads, their surface pattern is generally related to the orientation of the grain in the tetrad. The classical method of studying the mor­ phology of pollen has been through acetolysis. This method removes the protoplast, the intine and pollen coat substances, leaving only the exine. Studies of acetolyzed pollen provide in­ formation on the surface features. To obtain the best possible details, pollen grains should be mounted in an embedding medium of proper refractivity.The most commonly used mounting media have been glycerol, glycerol jelly and sili­ con oil. It is possible to get information on the struc­ ture of the wall at different levels by studying acetolyzed pollen through optical sectioning (careful focusing at different levels from the sur­ face towards the centre). Studies using micro­ tome sections provide a better understanding of exine elements. In recent years, electron mi­ croscopy (transmission and scanning) has pro­ vided a powerful technique for studying the finer details of the exine. distal pole

polar axis

equatorial axis

proximal pole

Fig. 2.10 Diagram of a tetrahedral tetrad of microspores to show the two poles and two axes (after Shivanna and Rangaswamy 1992).

38 Pollen Biology and Biotechnology

Apertures Apertures are the areas through which the pol­ len tube emerges during pollen germination (Rowley 1975). In the region of apertures, the exine is absent or very thin, thus facilitating tube emergence. Apertures are distinguished on the basis of shape, number and position (Fig. 2.11). They may be in the form of pores or in the form of furrows (colpi). The colpi are elongate boat­ shaped structures with more or less acute ends. Pollen grains with pores are referred to as porate and those with colpi as colpate. Pollen grains are called colporate when a furrow and a pore are combined in the aperture. In some pollen, the pores occur only in some furrows (generally located in half or one-third of the pollen). Such pollen grains are heterocolporate. The apertures, particularly the colpi, are also involved in accommodating changes in the vol­

ume of pollen as a result of hydration and desic­ cation (harmomegathic);the intine in desiccated pollen buckles inside in the apertural region bringing the two margins of the colpi together. It is suggested that this feature prevents exces­ sive desiccation of pollen. In hydrated pollen the margins expand, exposing the colpal surface. Phylogenetically the colpi are considered primi­ tive and the pores derived by contraction (Faegri and Iversen 1989). In pollen grains of many species the aper­ ture is covered with an isolated piece of exine called the operculum and such pollen grains are called operculate (Heslop-Harrison 1979c). In the porate type the operculum forms a disc (e.g. grasses) and in the colpate type a median strip (e.g. Potentilla). Edges of apertures may be simi­ lar to the rest of the exine (non-bordered) or may appear different to form a border (bordered). The border is generally in the form of a change in

Fig. 2.11 Diagram of the four types of pollen grains showing the arrangement of apertures in polar view (top row) and equatorial view (bottom row). Dotted lines show the apertures as seen In different focal planes. A, E. Trizonoporate. B, F. Trlzonocolpate. C, G.Trizonocolporate. D, H.Tripantoporate.

Pollen Morphology and Aeropalynology 39

size and/or density of exine sculpturing ele­ ments. The number (from one to forty or more) and arrangement of apertures varies between spe­ cies. The prefixes mono-, bi-, tri-, tetra-, penta-, hexa- and poly- are used to describe porate or colpate pollen. In a large number of dicotyledon­ ous species, three apertures (pores/colpi) cor­ responding to the three contact surfaces with other microspores of the tetrad are arranged equidistantly from each other along the equator. A great majority of monocotyledons show a single colpus or pore. When the apertures are arranged equidis­ tantly from each other along the equator or evenly distributed over the whole surface, the term zono- is used (e.g. zonopolyporate). When the pores are arranged irregularly, the term panto- is used (e.g. pantopolyporate). In some pollen grains, two or more colpi are combined into rings or a spiral surrounding the whole or a part of the grain (syncolpate).The colpi extend to different lengths towards the proximal and dis­ tal poles. The polar area is that part of the pollen which is above/below the level of colpi.The num­ ber, character and arrangement of apertures are very useful for identification of the pollen. Exine sculpture Exine sculpturing refers to the external surface features of the exine without reference to the internal structure of the exine (Faegri and Iversen 1989). Ornamentation of the exine is confined to the ektexine. The ektexine is basically a 3layerd structure with radially oriented columellae/baculae separating the outer roof, the tec­ tum and the lower foot layer (nexine 1). When the tectum covers all or most (ca 75%) of the surface, the pollen grain is called tectate (Fig. 2.12 A). When the tectum covers the pollen par­ tially (ca ggQ0^QQQc





M w i baculate


îq ïïd




tectate-perforate pilate

H D D DOC tectate-imperforate





Fig. 2.12 Diagram of intectate and tectate exine (A) and types of exine baculae as seen in sectional view in intectate grains (B).

Aerobiology is the study of biological particles such as pollen, fungal spores, dust mites, insect



scabrate granulate



various levels of the exine by careful studies of such optical sections of pollen by focusing at different levels. Erdtman (1956) called this type of study LO and OL analysis (from Latin Lux = light and obscuritas = darkness). LO for the sequence: light islands and dark channels (high focus) fol­ lowed by dark islands and light channels (low focus), and OL for the reverse sequence: dark islands and light channels followed by light is­ lands and dark channels. In recent years this method is hardly used in pollen analysis because simpler and more effective techniques for study­ ing exine structure are available.

>®©0**«& ® & ©°4>®@ @© 9 @© Ô ©Q ^® Vfi_ Ä a






Fig. 2.13 Basic sculpturing patterns of exine in surface view.

Pollen Morphology and Aeropalynology 41

debris and organic dusts present in the air (Hyde 1969,1972). Aeropalynology is the study of the release, dissemination, deposition and allergic effects of pollen grains and spores present in the air. It has been well established for more than a century that pollen grains are responsible for many allergic diseases, such as hay fever, asthma, allergic rhinitis and atopic dermatitis (Hyde 1969, Stanley and Linskens 1974, Knox 1979, 1993, Leuschner 1993, Agashe 1994). Allergy, also termed immediate hypersensitivity, is defined as an altered and accelerated reac­ tion of a person to a second or subsequent ex­ posures to a substance to which he/she has been sensitized during the first exposure. About 10-20% of the population is known to suffer al­ lergic disorders caused by bioparticles (Singh and Singh 1994).

Table 2.1 Some common plants in India in different seasons which cause pollen allergy (Source: Dr A.B. Singh, Centre for Biochemical Technology, Delhi) Spring (February-April)

Autumn (SeptemberOctober)


Cynodon dactylon Bothriochloa pertusa Dichanthium Cenchrus ciliaris annulatum Imperata cylindrica Heteropogon contortus Paspalum disticum Pennisetum typhoides Sorghum vulgare Poa annua Polygonum monspeliensis

Pollen grains induce allergic responses in sus­ ceptible individuals. Allergic pollen grains belong to three broad categories: grasses, weeds and trees. Rye-grass, timothy-grass, Kentucky grass and Bermuda grass are the major grasses. Rag­ weed (Ambrosia) is one of the important aller­ genic weeds, and birch, oak and hazel are some examples of tree allergens. Some of the com­ mon allergenic species in India are given in Table 2.1 (Singh and Rawat 2000). A characteristic feature of pollen allergy is its seasonal occur­ rence associated with the prevalence of pollen of that particular species in the atmosphere. Air­ borne pollen grains generally travel short dis­ tances; however, when they are blown into the upper strata of the atmosphere, pollen grains travel long distances before they are deposited. Meteorological factors, in particular temperature, precipitation, humidity and wind speed strongly influence airborne pollen counts. Pollen dispersal is facilitated by dry weather and higher wind ve­ locity. Pollen count is reduced after precipitation. Respiratory tracts, lungs, bronchi, skin and gas­ trointestinal tracts are commonly involved in al­ lergic reaction.

Cynodon dactylon Eragrostris tenella Phalaris minor Poa annua


Cannabis sativa

Allergic Response

Winter (NovemberJanuary)

Amaranthus spinosus Chenopodium Artemisia scoparia murale Parthenium Cassia occidentalis hysterophorus Ricinus Plantago major communis Suaeda fruticulosa Xanthium strumanium

Ageratum conyzoides Argemone mexicana Chenopodium album Asphodelus tenuifolius Ricinus communis


Alnus nitida Ailanthus excelsa Holoptelea integrifolia Prosopis juliflora Putranjiva roxburghii

Anogeissus pendula Eucalyptus sp. Prosopis juliflora Cedrus deodara

Cassia siamea Salvadora persica Mallotus phillipensis Cedrus deodara

In hay fever and asthma responses, pollen grain enters the nose and lands on the mucous membrane of the upper or lower respiratory tract. Pollen grain gets hydrated by the mucus secre­ tion and releases pollen allergens which pen­ etrate the mucous tissue. The allergens induce formation of antibodies (largely of IgE class) from B-lymphocytes (Fig. 2.14).The B-cells are stimu­ lated to proliferate clonally by cytokines (interleukins) which are produced by T-cells.

42 Pollen Biology and Biotechnology first exposure

subsequent exposure

allergenic pollen

release of allergens "

production of IgE

release of allergy mediators

mast cell Fig. 2.14 Schematic details of allergic response to pollen grains. First exposure to the allergen leads to the production of IgE antibodies which bind to the corresponding receptors on the surface of mast cells. On subsequent exposure the allergens released from the pollen bind to adjacent IgE molecules located on mast cells; this binding triggers release of various media­ tors of allergy such as histamines which induce allergic symptoms.

T-cell epitope, a specific region of the allergen molecule, is necessary for activation of T-cells. The antibodies produced from B-cells circulate through the serum and bind to the correspond­ ing receptors on the surface of the mast cells (Fig. 2.14) and basophils present in the connec­ tive tissue of the skin and other endothelial sys­ tems. There are about 105-6 receptor sites on each mast cell. On subsequent exposure, the allergens released from the pollen bind and cross-link specific IgE antibodies present on the surface of mast cells. This binding activates en­ zymes which release mediators such as hista­ mines, leukotrienes and prostaglandins from the mast cells (Fig. 2.14). These mediators induce allergic symptoms such as dilation of blood cap­ illaries, contraction of nasal and bronchial muscles, oversecretion of watery nasal fluid,

constriction of the nasal or bronchial passages, sneezing, itching and edema of the mucous membrane (Stanley and Linskens 1974, Chanda 1994). Elevated levels of IgE are detectable in the blood serum of allergic patients. The serum containing specific IgE can therefore be used as a probe for identification of allergens. Until recently, the relationship between pol­ len and asthma was not clearly understood as pollen grains are too large to reach the lower respiratory tract. Allergic asthma is the disease of airways and the allergens have to reach the lower respiratory tract to trigger allergic asthma. Recently allergen-containing microscopic par­ ticles released from bursting of pollen of grasses have been reported to occur in the atmosphere (Suphioglu et al. 1992). They provide a mecha­ nism by which pollen allergens may reach the

Pollen Morphology and Aeropalynology 43

lower respiratory tract (bronchi and lungs) and cause asthma (Ong et al. 1995a, b, 1996, Knox and Suphioglu 1996). These microscopic par­ ticles are < 5 pm and reach the respiratory tract. In grass pollen, allergen-containing microscopic particles have been shown to be intracellular starch granules/p-particles (Singh et al. 1991). These granules are released from pollen during rainfall when pollen grains rupture osmotically. Starch granules are readily detected in the at­ mosphere during the grass pollen season. On days following rainfall, a 50-fold increase in gran­ ule number per cubic metre of air has been re­ ported (Suphioglu et al. 1992).

Diagnostic Tests Several diagnostic tests, in vivo as well as in vitro, are available to screen allergenicity of dif­ ferent antigenic extracts. Among in-vivo tests, the skin test (scratch test and prick test) is the most commonly used; it is simple, convenient and highly specific (Freeman and Noon 1911, Dreborg 1989, Lin et al. 1993). Bronchial provo­ cation test has also been used for asthmatic patients. Among the in-vitro assays, Paper Radio Immuno Sorbent Test (PRIST), Radio Allergo Sorbent Test (RAST) (Wide et al. 1967) and Enzyme Linked Immuno Sorbent Assay (ELISA) (Engvall and Pearlman 1971) are important. These tests estimate the amount of IgE in the serum. In the RAST test, allergens placed on filter paper discs are treated with human serum followed by radiolabelled antihuman IgE. The amount of radioactivity bound to the disc is estimated. In the ELISA test, the radioactive label is replaced with an enzyme label. Dot immunoblotting is another method in which small drops of antigen are applied on nitrocellulose paper and processed according to ELISA pro­ cedures. This is very convenient for routine di­ agnostic application since it permits analysis of many samples on a single strip of paper. Another method termed FAST (Fluorescent Allergo Sorbent Test), which uses a fluorescent

substrate, was recently developed (see Singh and Malik 1992).

Allergens The pollen allergens are generally proteins or glycoproteins of molecular weight ranging from 5000 to 90,000 daltons. Using clinical and im­ munological approaches, a large number of al­ lergens have been identified in pollen grains of different species. Crude pollen extract contains many other components besides allergens. The methods of pollen collection and processing also affect the amount of allergens recovered (Stanley and Linskens 1974, Singh and Singh 1994). Variations in allergens have also been reported between pollen samples of different populations and also those collected in different years. Standardization of allergenic extracts is necessary for effective clinical trials and for im­ munotherapy. Recently the immunoprint/immunoblot tech­ nique has been extensively used to analyse al­ lergenic components in different pollen samples (Towbin et al. 1979). The extract separated through immunoelectrofocusing (IEF) is trans­ ferred from the gel to the nitrocellulose paper and the allergens are detected through specific IgE antibodies. In rye-grass pollen, 14 IgE bind­ ing proteins have been reported using sera of allergic patients (Ford and Baldo 1986). In Dactylis glomerata SDS-PAGE showed 13 al­ lergenic bands in the range of 14-17 kD (Ford et al. 1985). Similarly, of the 44 components of Cynodon dactylon, 17 are IgE binding proteins (Ford and Baldo 1986). Four allergenic compo­ nents have been detected from pollen of Parthenium hysterophorus (Subba Rao 1984). Allergenic components have also been identi­ fied in pollen of Artemisia (Jaggi and Gangal 1986). In Betula verrucosa a polypeptide of 17 kD has been identified as the major allergen (Ipsen and Lowenstein 1983). In recent years, monoclonal antibodies have been prepared for several major allergens (Westphal et al. 1988, Walsh et al. 1990, Valenta et al. 1992, Perez et al. 1990, Roberts et al. 1992) and are being used

44 Pollen Biology and Biotechnology

for identifying/localizing allergens (Singh et al. 1991, Grote et al. 1994, Knox and Suphioglu 1996). During the last 10 years several major aller­ gens (for example, pollen of ragweed, many members of Poaceae and white birch) have been cloned and their amino acid sequence deduced (see Singh et al. 1991, Swoboda et al. 1996, Ong et al. 1996, Smith et al. 1966, Knox and Suphioglu 1996). This has led to a better under­ standing of their primary structures and their biological functions. Lolp 1, a major allergen of rye-grass is expressed by a pollen specific gene (Griffith et al. 1991, 1993) and is an acidic gly­ coprotein with about 5% carbohydrate moiety (mol wt 35 kD). Deduced amino acid sequences of many of the allergens show high homology to a new class of conserved pathogenesis-related proteins. Rye-grass pollen allergens have been localized immunocytochemically using mono­ clonal antibodies and immunogold labelling (Knox and Suphioglu 1996). Lol p 1 proteins are localized in pollen cytoplasm and to a lesser extent in exine cavities (Singh et al. 1991, Taylor et al. 1993,1994) whereas Lol p 5 occurs largely in the starch granules and its molecular weight ranges from 28-32 kD. Ragweed pollen is one of the most impor­ tant allergens in the USA. Apart from the major allergenic proteins, Amb a 1 and Amb a 2, its

pollen contains many other allergenic proteins (Rogers et al. 1996). Several allergens from rag­ weed pollen have been characterized by pro­ tein sequencing and/or c-DNA cloning (see also Chapter 15). The c-DNAs of some of the aller­ gens have been expressed in E. coli and puri­ fied. Most of the recombinant allergens retain their IgE binding capacity. However, recombinant Cyn d 1 allergen required expression in yeast, a eukaryote, in order to restore IgE binding (Suphioglu et al. 1996).

Biological Standardization of Allergens This refers to the calibration of the potency of different allergenic preparations for their biologi­ cal activity (Aas 1978, Turkeltaub 1989). Intra­ dermal skin test and prick skin test have fre­ quently been used for standardization of the potency. One of the standard units followed by European workers is the HEP unit. An HEP unit is defined as the concentration of the extract that gives, in the skin prick test, a wheal diameter equivalent to 1 mg M of histamine hydrochlo­ ride (Dirksen et al. 1985). In the American method, intracutaneous skin tests are performed and comparison of extracts is made on their ability to induce a mean erythema diameter sum of 50 mm. (Turkeltaub 1989).This method is not related to histamine.

3 Pollen Viability and Vigour Pollen grains, following their release, are ex­ posed to the prevailing environmental conditions for varying periods before they land on the stigma. Depending on the conditions and the duration of exposure, the quality of pollen grains may be affected. The quality of pollen is as­ sessed on the basis of viability and vigour.

POLLEN VIABILITY Pollen viability refers to the ability of the pollen to perform its function of delivering male gametes to the embryo sac. The period for which pollen grains remain viable after they are shed varies greatly from species to species. On the basis of their longevity pollen grains of different species can be grouped into three categories (Harrington 1970, Barnabas and Kovacs 1997): 1 . Short-lived pollen: Pollen grains belonging to members of Poaceae, Cyperaceae, Alismataceae, Juncaceae, Commelinaceae and Asteraceae, which lose viability within a few days. In some species viability is lost in less than an hour. For example, under dry and warm conditions pollen grains of sor­ ghum (Fig 3.1), wheat and many other ce­ reals lose viability within 30 min (Fritz and Lukaszewski 1989, Lansac et al. 1994), and those of Compositae in about 3 h (Hoekstra and Bruinsma 1975a). 2. Pollen with medium lifespan: Pollen of a ma­ jority of families, such as Solanaceae,

Rutaceae, Cruciferae, Ranunculaceae, Liliaceae and Amaryllidaceae, fall within these extremes; they maintain viability for 1-3 months. 3. Long-lived pollen: Pollen of many Gymnosperms (Pinaceae and Gingkoaceae) and members of several angiosperm families, such as Leguminosae, Rosaceae, Anacardiaceae, Saxifragaceae and Arecaceae, maintain viability for over 6 months.

60 50 5 0s- 40 C

o "ro 30 c E 0 (D 20 V-







Time of desiccation (min) Fig. 3.1 Effect of desiccation on viability of pollen of Sor­ ghum assessed on the basis of in-vitro germination. Viability is lost within 30 min. (after Lansac et al. 1994).

46 Pollen Biology and Biotechnology

Comparative studies carried out on 2- and 3-celled pollen (Brewbaker 1959, Hoekstra and Bruinsma 1975b, Johri and Shivanna 1977, Hoekstra 1979) have demonstrated a broad cor­ relation between the cytology of pollen and their viability (see Table 1.2).Two-celled pollen grains in general retain viability for a longer period than 3 -celled pollen. Studies have shown that the rate of respiration in 3 -celled pollen (maintained un­ der high RH) is 2-3 times higher than 2-celled pollen (Hoekstra and Bruinsma 1975b). Although higher temperatures increased the rate of res­ piration in both 2- and 3-celled pollen, it was al­ ways higher in 3 -celled pollen than that in 2celled pollen (Fig. 3.2). These studies indicate that high respiratory activity may be a causative factor for rapid loss of viability of 3 -celled pollen (Hoekstra and Bruinsma 1975b); however, there is no direct evidence for such a suggestion. Rapid loss of viability in pollen grains of Poaceae is associated with their inability to with­ stand desiccation. Unlike pollen of other spe­ cies which are shed under desiccated condition (moisture level < 20%), pollen grains of Poaceae are shed under hydrated conditions (moisture

level ca 50%); loss of moisture is detrimental to their viability. The moisture content of the pollen is con­ ventionally expressed as the percentage of fresh weight (based on loss with oven-drying to con­ stant weight). Dumas and his associates (Dumas et al. 1985, Kerhoas et al. 1987) have used nuclear magnetic resonance (NMR) spectrom­ etry, which is non-destructive, and seems to be more accurate in determining the moisture level of pollen. Most of the water present in desic­ cated pollen is in bound form.

Tests for Viability For any experimental studies on pollen, assess­ ment of their viability is a prerequisite. Consid­ eration of pollen viability is also important in stud­ ies on pollen storage, reproductive biology and hybridization (Heslop-Harrison et al. 1984, Stone et al. 1995, Dafni and Firmage 2000). Standard­ ization of a simple, rapid and dependable test to assess pollen viability is important. A number of tests have been developed over the years to test pollen viability. Detailed protocols for commonly used tests are given in Shivanna and Rangaswamy (1992). Fruit- and seed-set



£ "c j0> o Q. U) J




£ n

ow o 0.1 10


20 25 30 temperature (°C)


Fig. 3.2 Respiration rate in 3-celled (A, B) and 2-celled (C, D) pollen at 97% humidity and different temperatures. A— Aster, B— Chrysanthemum, C— Nicotiana, D— Corylus (after Hoekstra and Bruinsma 1975b).

As viability refers to the ability of pollen to de­ liver functional gametes to the embryo sac, the most authentic test for viability would be to as­ sess the fertilization capacity of the pollen sample as measured by fruit- and seed-set fol­ lowing controlled pollination (Heslop-Harrison et al. 1984, Shivanna and Rangaswamy 1992). However, this test has many limitations for use as a routine test: (i) it is laborious and time-con­ suming; it may take many days or weeks before seed-set is assessed; (ii) many other factors such as stigma receptivity and incompatibility have to be taken into consideration to perform this test; (iii) seed-set is not an inevitable out­ come of fertilization as many post-fertilization factors associated with seed development may influence seed-set; (iv) this test cannot be used in apomictic species; (v) it can be used only

Pollen Viability and Vigour 47

during the flowering period of the species; (vi) it can be used more as a qualitative than a quan­ titative test, particularly in systems with fewer ovules, as germination of a limited number of pollen is enough to induce full seed-set. There­ fore, assessment of pollen viability through fruitand seed-set is not practicable as a routine test, although it can be used to confirm the results of other tests.

Pollen germination and pollen tube growth in the pistil As an alternative to fruit- and seed-set, some attempts have been made to assess pollen vi­ ability by studying pollen germination and pol­ len tube growth in the pistil following controlled pollinations. In Brassica oleracea, for example, pollen samples which produce 70 pollen tubes in the style are considered fully viable (Ockendon 1974). Although this method markedly reduces the time taken compared to the fruit- and seedset method, it has most of the other limitations associated with fruit- and seed-set. Also, it is not always feasible to quantify the number of pollen tubes growing in the style.

Non-vital stains and other tests of limited use Because of the limitations of the above meth­ ods, many alternative methods which are simple, convenient and rapid have been developed over the years. Many of the staining tests using non­ vital stains such as iodine in potassium iodide, aniline blue in lactophenol, acetocarmine, acid fuchsin and Alexander’s stain (Alexander 1980), essentially assess the presence of contents in the pollen; they are satisfactory in assessing pollen sterility but are not dependable for test­ ing viability (see Heslop-Harrison et al. 1984). Some non-permeating stains such as Evans blue and phenosafranin (Widholm 1972) which do not enter plasma membranes of living cells but stain the cytoplasm of dead cells are report­ edly suitable for assessing pollen viability. How­ ever, studies using these tests have been lim­ ited so far. It would be worthwhile to extend them

to a larger number of species to assess their general applicability. Inorganic acid tests (Koul and Paliwal 1961, see Shivanna and Johri 1985) based on burst­ ing of viable pollen in inorganic acids and for­ mation of instant pollen tubes (Linskens and Mulleneers 1967, Maheshwari and Mahadevan 1978) from hydrated pollen, although very simple and rapid, are largely not used because data for establishing their correlation with true viabil­ ity are lacking. Similarly, estimation of ATP con­ tent (assayed by luciferase-luciferin method), shown to correlate with germinability in many conifers (Ching et al. 1975), has not been much used either, probably because of the time and expense involved in this test. A cytochemical test, termed the benzidine test, which is based on the oxidation of benzi­ dine by peroxidase in the presence of hydrogen peroxide, has been used in the past for assess­ ing viability of pollen of many species (King 1960). In recent years, however, this test has not been used because of benzidine toxicity and also the availability of better and effective tests.

Tétrazolium test The tétrazolium test is one of the most widely used, especially by early investigators (see Stanley and Linskens 1974). This test is based on reduction of soluble colourless tétrazolium salt to reddish insoluble formazan in the pres­ ence of dehydrogenase. Following incubation of pollen grains in tétrazolium solution for 30-60 min, pollen grains which take a reddish colour are scored as viable (Fig. 3.3A). Although 2,3,5triphenyl tétrazolium chloride is the most com­ monly used salt, many other salts such as nitroblue tétrazolium (which is specific to suc­

cinic dehydrogenase) (Hauser and Morrison 1964) and tétrazolium red (Aslam et al. 1964) have also been used. Some investigators have reported satisfac­ tory results with this test in assessing pollen vi­ ability in several species (Hauser and Morrison 1964, Norton 1966, Collins et al. 1973).

48 Pollen Biology and Biotechnology

• •



Fig. 3.3 Tests for pollen viability. A.Tétrazolium test; viable pollen grains are darkly stained. B. Fluorescein diacetate test (after Shivanna and Rangaswamy 1992).

However, many investigators have reported false positive responses with tétrazolium test (Oberle and Watson 1953). Often, results of the tétrazo­ lium test did not correlate with seed-set data (Barrow 1983) or the in-vitro germination test (see Heslop-Harrison et al. 1984). For example, in Simmondsia over 95% of stored pollen devel­ oped colour reaction in tétrazolium solution even in samples which did not show in-vitro germina­ tion (Beasley and Yermanos 1976). Similarly, over 90% of the pollen of Helleborus niger sub­ jected to DMSO or heat treatment, which failed to respond to germination or FDA tests, were positive for the tétrazolium test (Fig. 3.4) (HeslopHarrison et al. 1984). Another limitation of the tét­ razolium test is that the colouration of respond­ ing pollen shows a gradation from very light to dark red with the resuit that the cut-off point for scoring viable pollen becomes subjective. Hence the tétrazolium test has not been popular in re­ cent years.


□ germination □ FDA

ED tétrazolium






0 Fresh

DMSO Treatment


Fig. 3.4 Comparison of in-vitro germination, FDA and tétra­ zolium tests on fresh, DMSO-treated and heat-treated pol­ len of Helleborus niger: A good correlation was found be­ tween all the tests with fresh pollen. Tétrazolium test did not reflect viability in DMSO- and heat-treated pollen (after Heslop-Harrison et al. 1984).

Pollen Viability and Vigour 49

In-vitro germination test The most commonly used and acceptable test for assessing pollen viability is the in-vitro ger­ mination test. It is rapid, simple and the results of in-vitro germination generally correlate with seed-set data (Akihama et al. 1978, Janssen and Hermsen 1980). However, this correlation de­ pends on optimization of the medium and other cultural conditions to induce germinability in most of the viable pollen. In suboptimal medium this test gives false negative results. Further, many of the stored pollen samples which fail to germi­ nate in vitro are often found to be capable of inducing fruit- and seed-set following pollination (Visser 1955, King 1963,1965, Ghatnekarand Kulkarni 1978). A major limitation of this test is lack of optimal germination medium for pollen of many species, particularly the 3-celled pollen species. Fluorescein diacetate test The fluorescein diacetate (FDA) test, often re­ ferred to as the fluorochromatic reaction (FCR)

test, was introduced by Heslop-Harrison and Heslop-Harrison (1970) as a test for pollen vi­ ability. This has been the most commonly used test in recent years. The FDA test assesses two properties of the pollen: (i) integrity of the plasma membrane of the vegetative cell and (ii) pres­ ence of active esterases in the pollen cytoplasm. Non-polar, non-fluorescing FDA passes freely through the pollen membrane and enters the pollen cytoplasm. Hydrolysis of FDA by the ac­ tivity of esterases results in fluorescein which is fluorescent. Since fluorescein (a polar sub­ stance) does not pass through the intact plasma membrane as readily as FDA, it accumulates in the pollen cytoplasm (Fig. 3.5). Such pollen grains show bright fluorescence when observed under a fluorescence microscope (Fig. 3 .3 B) (blue light excitation). Pollen grains that do not have intact plasma membrane allow fluorescein to move out readily (Fig. 3.5) and thus result in uniform background fluorescence. Likewise, if there are no active esterases in the pollen cyto­ plasm, fluorescein is not formed and hence pol­ len grains do not fluoresce.

intact functional — plasma — membrane disrupted / nonfunctional nonviable pollen cytoplasm

viable pollen cytoplasm



fluorescein diacetate


)Xl fluorescein



Fig. 3.5 Basis of fluorescein diacetate (FDA) test. FDA readily enters the pollen cytoplasm; in the presence of active es­ terases, FDA is hydrolyzed to release fluorescein, a fluorescent substance, and acetate. As fluorescein cannot readily pass through the intact membrane, it accumulates inside the cytoplasm in pollen grains with intact plasma membrane (left) and thus such pollen grains show bright fluorescence (as seen in Fig. 3.3 B). Fluorescein readily moves out of pollen cytoplasm with disrupted plasma membrane (right); such pollen preparations show a general background fluorescence and not the bright pollen fluorescence.

50 Pollen Biology and Biotechnology


I • c o 03

c E I— 0) 50 0 1 i o < Q








time (min) Fig. 3.6 Effect of controlled hydration on FDA and in-vitro germination tests of desiccated pollen of Carex ovalis. Des­ iccated pollen grains neither responded to FDA test nor showed in-vitro germination. Controlled hydration dramati­ cally restored membrane integrity and germinability (after Shivanna and Heslop-Harrison 1981).

The FDA test has proven satisfactory in as­ sessing pollen viability in a number of species (Shivanna and Heslop-Harrison 1981, HeslopHarrison et al. 1984, Knox et al. 1986, Shivanna and Cresti 1989, Shivanna et al. 1991a, b, Rao et al. 1992, 1995, Sedgley and Harbard 1993). It reportedly has wider applicability and better resolution than other prevailing tests for assess­ ing pollen viability of cotton (Gwyn and Stelly 1989). In species in which the medium used for invitro germination of pollen is optimal, a close correlation exists between the FDA test and invitro germination test (Shivanna and HeslopHarrison 1981, Heslop-Harrison et al. 1984). In

the absence of optimal medium, the FDA test gives a better index of viability than in-vitro ger­ mination (Shivanna et al. 1991a).The FDA test reflected the fertilizing ability of pollen grains exposed to high humidity (> 90% RH) and high temperature stresses (up to 60°C) better than the in-vitro germination test (Shivanna et al. 1991b). To obtain valid results with the FDA test some important precautions have to be taken. For as­ sessing viability of dry and desiccated pollen samples, they have to be exposed to controlled hydration (by maintaining them under high hu­ midity for about 30 min) before testing for viabil­ ity (Fig. 3.6); such pollen may not respond if tested directly (Shivanna and Heslop-Harrison 1981). The FDA test may not reflect fertilizing ability in pollen samples subjected to prolonged exposure to very high temperature (75°C) (Rao et al. 1995). In pollen samples of Brassica sub­ jected to 75°C, the extent of retention of pollen fluorescence gave a better indication of pollen viability than initial fluorescence. Pollen samples which induced seed-set retained fluorescence even after 2 h, while those ineffective in induc­ ing seed-set (treated at 75°C for 24 h) lost fluo­ rescence within 60 min due to leakage of fluo­ rescein.

Causes for Loss of Viability Until recently there were hardly any studies to understand the causative factors for the loss of viability. It was suggested that deficiency of res­ piratory substrates and/or inactivation of en­ zymes and growth hormones are responsible for loss of viability (Stanley and Linskens 1974). As the metabolic activity of the pollen contin­ ues, though at a very reduced rate, even after shedding (Wilson etal. 1979), endogenous sub­ strates are expected to be used up gradually and result in loss of viability. However, there are no clear evidences to show that loss of respira­ tory substrates is the primary cause for loss of viability, particularly in short-lived pollen. Pollen grains of cereals, which are short-lived, contain abundant reserve metabolites even after losing

Pollen Viability and Vigour 51

viability (Hoekstra and Bruinsma 1980). Loss of respiratory substrates, however, may be a fac­ tor in the loss of viability during long-term stor­ age. Similarly, there are no direct data to indi­ cate that inactivation of enzymes or deficiency of growth hormones are the causes for the loss of viability. Although a few investigators have reported changes in amino acid composition of stored pollen, there has been no clear correla­ tion between the loss of viability and changes in amino acid content (Stanley 1971, Linskens and Pfahler 1973, Dashekand Harwood 1974). Loss of membrane integrity Many studies carried out during the last 20 years have highlighted the role of plasma membrane in maintaining pollen viability (Heslop-Harrison 1979a, Shivanna and Heslop-Harrison 1981, Heslop-Harrison et al. 1984, Jain and Shivanna 1987a,b, 1989). Irreversible loss of membrane integrity seems to be the primary cause for the loss of pollen viability. A close correlation has been established between membrane integrity and germinability of pollen of a number of spe­ cies (Fig. 3.6).

Pollen grains of many species subjected to desiccation for different periods showed (Shivanna and Heslop-Harrison 1981) a marked variation in their ability to withstand desiccation. Pollen of several species lost membrane integ­ rity following 1-4 h desiccation (Fig.3.7A); such pollen grains invariably failed to germinate in vitro. However, pollen of some species such as Cytisus retained membrane integrity as well as germinability even after 24 h of desiccation (Fig. 3.7B). Pollen of some of the desiccation-sensi­ tive species such as Iris, Eleocharis and Lonicera, recovered membrane integrity as well as in-vitro germinability when desiccated pollen was subjected to controlled hydration before test­ ing (Fig. 3.8A, B).Thus, controlled hydration pro­ vides suitable conditions for restoration of mem­ brane integrity. This accords with many studies reporting favourable effects of controlled hydra­ tion on in-vitro germination (see Chapter 4). However, desiccated pollen grains of Secale failed to restore membrane integrity even after controlled hydration (Fig. 3.8C, D). These observations confirm that pollen grains of cereals in general are highly susceptible to




I c


o 03







£ L-








O c < 0 Q.

O c £to a5

18 14 10-




® 0) ? ! 2.0_£"

i !


= ■§ 0.6 w O


a) JC

0 .2 f ­


Sr S 0.04



0) £ 0 Q

o c as

TO >% E

CD O 40



2 Q.

O .y . O


v .v .y


/M ,

S2 product pollen exine S1 product

S1 precursor B Fig. 9.3 Tapetal origin of S-allele products in sporophytic systems and their incorporation into pollen exine (modified from Pandey 1979).

144 Pollen Biology and Biotechnology

genes is required for SI. In members of Ranunculaceae and Chenopodiaceae, SI is re­ ported to be controlled by four multiallelic genes that are linked in their inheritance (Lundqvist 1975, 1990, Osterbye 1975, Larsen 1977). Ac­ cording to Lundqvist (1975) SI controlled by many genes is the primitive one and that con­ trolled by a single gene the derived condition. Even in members of Cruciferae (Eruca sa­ liva, Raphanus sativus and Brassicacampestris) SI has been reported to be determined by mul­ tiple genes (Verma et al. 1977, Lewis 1979). Lewis and associates (1988) and Zuberi and Lewis (1988) explained that in R. sativus and B. campestris, SI is controlled by two linked genes, S and G; the former has sporophytic control and the latter gametophytic control in the pollen. Matching of alleles at both S and G loci is re­ quired for SI.These researchers consider the G locus the ancestor of the S-gene in the gameto­ phytic system that has been retained in the sporophytic system. Determination of the genetics of SI, in par­ ticular of SSI, is time consuming and laborious; it involves a large number of controlled pollina­ tions and analyses of the progeny often extend­ ing to many generations (Wallace 1979).There­ fore, only a few species have been genetically analysed. However, studies of other correlated features (Table 9.1) give an indication of the genetic basis of SI. The number of alleles in a population is po­ tentially large. Over 40 S-alleles have been re­ ported in populations of Oenothera organensis (Emerson 1939), Papaver rhoeas (Lawrence et al. 1993) and Lolium perenne (Fearon et al. 1994). In Sinapis arvensis, Stevens and Kay (1989) reported 35 different S-alleles from 35 plants raised from seeds collected randomly from a population. In Brassica oleracea (Ockendon 1975), 49 alleles have been identi­ fied in a wide range of cultivars.

Cytology Cytological details of pollen inhibition have been investigated in a number of species. In SSI inhi­

bition is strictly confined to the stigma surface. Self-pollen grains either fail to germinate or the small tubes that emerge, at most, penetrate the cuticle (Kanno and Hinata 1969, Dickinson and Lewis 1973a, b). Inhibition of pollen is generally associated with the deposition of callose in the germ-pore or at the tube tip (Figs. 9.4, 9.5). Interestingly, stigmatic papillae in contact with incompatible pollen, also develop a lenticular plug of callose at the tip, between the cell wall and the plasma membrane (Dickinson and Lewis 1973a, b, Heslop-Harrison et al. 1974, Howlett et al. 1975, Shivanna et al. 1978a, Kerhoas et al. 1983). This reaction is termed the rejection reaction (Fig. 9.4A, B, 9.5). No such plugs de­ velop following compatible pollination.The depo­ sition of callose plug in the papillae is rapid and is often apparent within 10 min of pollination. Although it is a clear manifestation of self-polli­ nation in SSI systems, it does not seem to play a direct role in pollen tube inhibition. Inhibition of callose deposition by treating the stigmas with 2-deoxy-D-glucose before self-pollination (Singh and Paolillo 1989) showed that callose deposi­ tion in the stigmatic papillae is not essential for pollen inhibition (see also Sulaman et al. 1997). Rejection reaction on the stigmatic papilla has been used as an assay to analyse the loca­ tion of pollen components involved in SSI in Raphanus and Iberis (Dickinson and Lewis 1973a, Heslop-Harrison et al. 1974). Deposition of exine material (isolated through high-speed centrifugation of pollen or leached onto an aga­ rose film) from the self-pollen onto the stigmatic papillae stimulated development of the rejection reaction; the pollen exine material from cross­ pollen was not effective. Thus the exine compo­ nents are involved in at least one phenotypic manifestation of SI, development of the rejec­ tion reaction. In gametophytic species in general, pollen germination is not affected but pollen tubes are inhibited in the style (de Nettancourt 1977). How­ ever there are variations. In members of Papaveraceae, the zone of inhibition is gener­ ally the stigma. In some species such as

Self-incompatibility 145

Fig. 9.4 Pollen Inhibition in self-pollinated pistils. Fluorescence micrographs of the stigma 6 h after self-pollination stained with decolourized aniline blue. A, B. Brassica campestris. Most of the papillae show rejection reaction (development of callose plug at the tip). A few of the papillae are magnified in B. C-E. Saccharum bengalensis. Pollen grains are inhibited before germination (C), after germination but before pollen tube entry (D), or after growing a short distance In the stigma (E); callose has accumulated at the germ pore and pollen tubes. E. Lycopersicon peruvianum. Pollen tube (pt) is swollen at the tip and has burst in the transmitting tissue of the style (A-E—after Shivanna 1982, F—after de Nettancourt et al. 1974).

Oenothera (Dickinson and Lawson 1975a) and members of Commelinaceae (Owens 1981), pollen tubes are inhibited in the tissues of the stigma itself. In many species of grasses (Heslop-Harrison et al. 1974, Sastri and Shivanna 1979) the zone of inhibition is variable often within the same species (Fig. 9.4C-E). In­ hibition may take place at the levels of pollen germination, pollen tube entry into the papillae, and pollen tube growth at different levels in the stigmatic tissue. In some species, inhibition is delayed until the tubes reach the main axis of the pistil (Thomas and Murray 1975). Pollen tube inhibition is associated with deposition of an excessive amount of callose in the tube, par­ ticularly at the tip (Fig. 9.4C-E).

The tips of incompatible tubes often show swelling and/or bursting in the stylar region. In Lycopersicon peruvianum (de Nettancourt et al. 1974), compatible tubes have typical two-layered wall—an outer pectocellulosic wall of loose fibrils and an inner homogeneous callose wall. Incom­ patible tubes display similar wall structure dur­ ing initial growth. As they grow down one-third of the style, the inner layer gradually becomes thin and numerous particles (ca 0.2 |jm diam­ eter) accumulate in the tube cytoplasm. Also, the endoplasmic reticulum in the incompatible tubes is altered into a whorl of concentric lay­ ers. Eventually the inner wall disappears com­ pletely and the tube tip bursts releasing these particles into the intercellular spaces of the

146 Pollen Biology and Biotechnology

pollen grain callose « papilla pollen tube pellicle cuticle

Fig. 9.5 Diagram of early events of poilen-pistii interaction in sporophytic SI systems. A. Cross-pollination. B. Self-pollina­ tion. Note development of callose plug (rejection reaction) in the stigmatic papilla following self-pollination (after Shivanna 1982).

transmitting tissue (Fig. 9.4F). On the basis of such evidence de Nettancourt and colleagues (1975) suggested that SI is an active process resulting in the destruction of the pollen tube wall, and inhibition of protein synthesis a result of al­ tered endoplasmic reticulum. Although excessive deposition of callose in the pollen tube is a characteristic feature of self­ pollen tubes (Fig. 9.4C-F), detailed studies of Shivanna and co-authors (1978b, 1982) on SI response of grasses have shown that callose deposition is not the primary event and there­ fore not the cause of inhibition. Deposition of pectic material at the tube tip is the earliest de­ viation observed in selfed tubes. Callose is ini­ tially deposited behind the tip region and later may extend to the tip.

al. 1979, Roberts et al. 1980). The extent of ad­ hesion and hydration (Fig. 9.6) was much less in self-pollen than in cross-pollen. There are many differences in the metabo­ lism between self- and cross-pollinated pistils. Detailed studies have been carried out on many enzymes, in particular peroxidases (Pandey 1967, Bredemeijer 1974, 1982) and esterases (Pandey 1973), in self- and cross-pollinated pis­ tils mainly to analyse their role in the inhibition of self-pollen tubes. Changes have also been shown in the synthesis of RNA and proteins in the style and ovary in response to self- and cross-pollination (van der Donk 1974, Deurenberg 1976,1977b). However, to date no evidence has been adduced that any of these differences is directly involved in pollen tube in­ hibition (see Shivanna and Johri 1985). One of the effective approaches to studying the basis of SI recognition and inhibition is through the use of an in-vitro assay in which self­ pollen grains are selectively recognized and inhibited (see Shivanna and Johri 1985). Many such assays have been reported from time to


cross-poller self-pollen

80 -o






a> ôCL sP 0s*



Physiological and Biochemical Studies A number of physiological and biochemical dif­ ferences in the pollen and/or pistil between selfand cross-pollinations have been reported in both GSI and SSI systems. In Brassica, differ­ ences in pollen adhesion and hydration, and mobility of pollen wall components between selfand cross-pollen have been reported (Stead et





time Fig. 9.6 Differences in pollen hydration following self- and cross-pollinations in Brassica oleracea. Pollen hydration is expressed as % of the total number of pollen grains that have become spheroidal (after Roberts et al. 1980).

Self-incompatibility 147

time using crude/dialyzed extracts (see Jackson and Linskens 1990). Although Ferrari and Wallace (1975) reported selective inhibition of in-vitro pollen germination by self-stigma leachate in Brassica oleracea, subsequent studies failed to record such a differential effect (Roberts et al. 1983). One of the major difficulties in establishing such an assay in Brassica was lack of a suitable medium to achieve satisfactory in-vitro germination and tube growth. Later many media were formulated to achieve good in-vitro germination (Roberts et al. 1983, Hodgkin and Lyon 1983) as well as tube growth (Shivanna and Sawhney 1995). Using one of these media Singh and Paolillo (1989) reported differential effects of stigmatic eluates on self- and cross­ pollen. The progress made through use of in-vitro assay has been more substantial in gameto­ phytic systems. In Petunia hybrida (Sharma and Shivanna 1982, 1983a) as well as Nicotiana alata (Sharma 1986), both crude as well as dialyzed extracts from unpollinated pistils inhibited in-vitro germination and tube growth of self-pol­ len but not of cross-pollen (Fig. 9.7A, B, 9.8A-

C). Pistil extract from the bud, which permits in­ compatible pollen to grow through, did not show inhibition of self-pollen. Incorporation of a transcription inhibitor (actinomycin D/cordycepin) in the culture medium (containing pistil extract) was effective in over­ coming inhibition of self-pollen (Fig. 9.8D) (Shivanna and Sharma 1984). These studies clearly established that (i) the S-allele compo­ nents are synthesized in the pistil before polli­ nation and (ii) inhibition of self-pollen is depen­ dent on fresh transcription in the pollen. Subse­ quently an in-vitro assay was established by in­ corporating isolated S-glycoproteins (Jahnen et al. 1989). In many animal systems and lower plants (see Marx 1978, Shapiro and Eddy 1980), it has long been known that gamete recognition is es­ tablished as a result of complementation be­ tween surface saccharides of one gamete and saccharide-specific receptors (lectins) on the other gamete.This is the result of in vitro experi­ ments in which gamete recognition is inhibited when the recognition molecules are blocked by specific sugars or lectins.

350 100




300 control

B 801 c o w


□ ? 200

60 ■

40 100


Ea 5 200

O ) c © g 150 3 M


100 50


+ cross­ extract

+ self­ extract


time (h)

Fig. 9.7 In-vitro assay for self-incompatibility inhibition in Petunia hybrida. A. Effects of self- and cross-pistil extracts on pollen germination and pollen tube growth. Cross-extract had no effect while self-extract inhibited both pollen germination and tube growth. B. Temporal details of pollen tube growth in control and presence of self- and cross-pistil extract (after Sharma 1986).

148 Pollen Biology and Biotechnology

Fig. 9.8 In-vitro assay for self-incompatibility In Petunia hybrida. Fluorescence micrographs of pollen cultures grown in vitro in the presence of pistil extract. A. Control medium (no pistil extract), B. Cross-pistil extract, C. Self-pistil extract, D. Self-pistil extract + actinomycln D. Actinomycin D overcomes the inhibition of self-pistil extract (after Sharma 1986)

Using in vitro pollen germination assay, Sharma and Shivanna (1983a) and Shivanna and Sharma (1984) carried out experiments in Petunia comparable to those in animal systems. Pollen grains were treated with lectins or sugars before culturing in a medium containing pistil extract. Inhibition of self-pollen was overcome when pollen grains were treated with glucose or lactose but not with lectins. Similarly, incorporation of specific lectins into the medium containing pistil extract was also effective in overcoming inhibition of self-pollen. However, neither the sugars nor the lectins showed Sallele specificity. Although these results indicate that recognition of self-pollen is established as a result of complementation between lectin-like components of the pollen and specific sugar moiety of the pistil, presumably of S-allele specific glycoproteins, recent studies (see page 149-154) on characterization of S-allele components have not indicated a role for carbohydrates of S-allele glycoproteins in pollen recognition or inhibition. In Papaver rhoeas also a consistent in vitro assay has been established and used effectively to understand the details of inhibition (FranklinTong et al. 1988,1989, Foote et al. 1994). Addi­ tion of crude stigma extract and, more impor­

tantly, of purified stigmatic S-protein or recom­ binant S-gene product, was effective in inhibi­ tion of tube growth of self-pollen. As in Petunia (Shivanna and Sharma 1984), in Papaver also incorporation of actinomycin D in the in-vitro assay could overcome inhibition of self-pollen. These results confirm that SI inhibition requires de novo pollen gene expression (Franklin-Tong etal. 1990).

Temporal Expression of SI In both GSI and SSI, expression of the self-in­ compatibility response is developmentally regu­ lated. The SI is weak or even absent in imma­ ture flower buds, while mature flowers are strongly incompatible (Shivanna and Johri 1985). For example, in Petunia as well as Brassica self­ pollination of buds 2-4 days before anthesis re­ sults in fruit- and seed-set. This has been ex­ plained on the basis that immature buds do not contain S-allele products or only a low concen­ tration; S-allele components reach effective con­ centration only towards maturity. This knowledge of bud compatibility has been effectively used to produce homozygous lines for breeding pur­ poses. More importantly, bud compatibility has proved invaluable in isolation and characteriza­ tion of S-genes.

Self-incompatibility 149

Operation of SI SI involves highly specific cellular and molecu­ lar recognition. For effective operation of SI, the following are the essential requirements: • Production of S-allele specific products in the pistil and the pollen. • Interaction of S-allele specific products of the pistil with those of the pollen at some stage of pollen-pistil interaction to establish pollen recognition. • Inhibition of incompatible pollen following pollen recognition. An understanding of the operation of SI there­ fore involves isolation and characterization of S­ alleles and their products in the pistil and pol­ len, and elucidation of the basis of pollen recog­ nition and the mechanism of pollen inhibition. Remarkable progress has been made during the last two decades in our understanding of the details of S-genes and their products and, to a lesser extent, the details of pollen inhibition. The basis of pollen recognition is yet to be clearly elucidated, however.

Characterization of S-allele Products and S-alleles in Pistil Sporophytic systems Initial progress in understanding S-allele prod­ ucts was made by Nasrallah and associates in Brassica oleracea using immunodiffusion and electrophoretic studies (Nasrallah and Wallace 1967, Nasrallah et al. 1970,1972).They showed that each S-allele produces a specific protein and each S-allele specific protein has a differ­ ent electrophoretic mobility; each could there­ fore be localized to a specific band in the gel. These proteins were heritable, as evidenced by the presence in the heterozygous plant, of both the S-allele specific bands of the parents. Seg­ regation of S-allele specific bands in F2progeny correlated precisely with segregation of the Sallele phenotype. Acquisition of self-incompat­ ibility in the pistil coincided with the appearance of S-allele-specific proteins. In B. oleracea for

example, S-allele-specific proteins in the stigma increased in concentration with the onset of SI about 2 days prior to anthesis; immature buds about 5 days before anthesis, which are self­ compatible, contained only trace levels of S-al­ lele proteins (Nasrallah 1974, Nasrallah et al. 1985, Roberts et al. 1979). Subsequently Hinata and associates (Nishio and Hinata 1977, 1978, 1979, 1980, 1982, Hinata and Nishio 1981) demonstrated the pres­ ence of S-allele specific band(s) following iso­ electric focusing in B. oleracea as well as B. campestris. These proteins were shown to be glycoproteins and many were purified from stigma extracts by gel filtration followed by affin­ ity chromatography on concanavalin Asepharose column (Nishio and Hinata 1979). All S-allele glycoproteins showed binding to a lec­ tin, concanavalin A (Nishio and Hinata 1980, Hinata and Nishio 1981). The molecular weight of all the S-allele glycoproteins in Brassica has been estimated to be around 55 kD; they usu­ ally have high isoelectric points (pi 10.3-11.1). The amino acid sequences of many S-glycoproteins have been determined (Nasrallah et al. 1987, Takayama et al. 1987).These studies apart from establishing S-allele glycoprotein-SI phe­ notype relationship, indicated that the gene(s) encoding these glycoproteins represent either the S-genes or sequences tightly linked to Sgenes. All S-allele specific glycoproteins of Brassica have 12 conserved cysteine residues located in the C-terminal region of the proteins (Nasrallah et al. 1987). These glycoproteins show 13 po­ tential N-glycosylation sites; of these. 4 are con­ served among all the alleles (see Dickinson et al. 1992). Analyses of carbohydrate chains of three S-allele specific glycoproteins from B. campestris have shown that the major N-glycosidic oligosaccharide chains A and B are identi­ cal and are commonly found in other plant gly­ coproteins (Takayama et al. 1986), rather than being specific to S-gene products of Brassica. These studies indicate that oligosaccharide

150 Pollen Biology and Biotechnology

moiety may not play a role in S-allele specificity (Takayama et al. 1986). Nevertheless, glycosylation seems to be required for SI; treat­ ment of stigma with tunicamycin, an inhibitor of glycosylation, overcomes inhibition of self-pol­ len and results in normal germination and tube growth (Sarker et al. 1988). Recent studies using molecular approaches have shown that the S-gene in Brassica is a complex multigene family and contains a se­ creted S-locus specific glycoprotein gene (SLG) and an S-locus receptor kinase gene (SRK) (Lalonde et al. 1989, Nasrallah and Nasrallah 1993, Hinata et al. 1993, Sims 1993, Franklin et al. 1995). The SLG protein is around 57 kD and the SRK protein is around 120 kD. Besides these, another group of the S-gene family, S-locus re­ lated genes (SLR-1 and SLR-2), has also been identified. S-locus-specific glycoprotein gene (SLG): Nasrallah et al. (1985) were the first to isolate a cDNA clone that contained sequences encoding part of the S-glycoprotein corresponding to S6 allele of B. oleracea by differential colony hybridization. Since then cDNA and genomic clones have been isolated for many other S­ alleles (Chen and Nasrallah 1990, Toriyama et al. 1991, Nasrallah et al. 1987, 1988, Trick and Flavell 1989, Nasrallah and Nasrallah 1993, Wheeler et al 2001). Analysis of RFLPs from parents, F, and F2populations have shown exact co-segregation of RFLP bands with specific S­ alleles, providing evidence for the identification of S-allele specific genes. Studies using Northern hybridization and in-situ hybridization (Nasrallah et al. 1985,1988, Kandasamy et al. 1989) have revealed that the S-gene is highly


expressed in the stigmatic papillae, the site of pollen inhibition, and its glycosylation products accumulate in high levels in the wall of the papillae. S-allele specific genes can be divided into two classes based on intensity of incompatibility (Nasrallah et al. 1991). Class I alleles exhibit a strong SI reaction and Class II alleles show a weaker SI and are recessive in pollen. Identification and characterization of S-receptor kinase gene (SRK): Nucleotide sequencing of the flanking regions of a number of SLG ho­ mologous genomic clones from B. oleracea re­ sulted in identification of the SRK gene (Stein et al. 1991). Studies on the structural details of the SRK gene suggest that SRK protein is an inte­ gral membrane protein with an external SLGlike receptor domain (which shares extensive sequence similarity with SLG), a transmembrane domain and a cytoplasmic kinase domain (Fig 9.9).The SLG-like domain of SRK6shares 94% nucleotide sequence identity with SLG 6 (Nasrallah et al. 1991). SRK2and SLG2similarly share over 90% identity. The identity between SLG2and SLG6and between SRK2and SRK6is only about 68% (Stein et al. 1991).The sequence of the cytoplasmic domain of SRK is very simi­ lar to the serine/threonine kinase. SRK exhibits the same developmental and spatial pattern of expression as the SLG gene (Stein et al. 1991). Analysis of RFLPs in F2 populations segregat­ ing for different SI genotypes has shown that SLG and SRK genes are genetically inseparable from each other and from the SI phenotype; they behave as components of a single locus (Fig. 9.9). On the basis of this data Nasrallah and associates (Nasrallah and Nasrallah 1993)



SLG-like | kinase domain domain transmembrane domain Fig. 9.9 Diagram of SLG and SRK genes in S locus of Brassica (after Nasrallah and Nasrallah 1993).

Self-incompatibility 151

adapted the term ‘S-haplotype’ instead o f ‘Saliele’ for SI genotypes. Isolation and characterization of S-locus related genes: The pattern of expression of SLR genes (SLR-1 and SLR-2), both spatially and tempo­ rally, is identical to the SLG gene; their expres­ sion is also confined to stigmatic papillae and the products reach maximum accumulation to­ wards maturity. However, unlike the SLG gene, SLR genes are not linked to the S-locus and therefore do not contribute to S-allele specific­ ity. Although SLR-1 and SLR-2 loci are distinct from SLG, SLR-1 and SLR-2 are linked (Boyes et al. 1997). Analysis of SLR-1 and SLR-2 genes from different Brassica genotypes revealed an extremely low level of allelic variation (99-100% homology) (Trick 1990, Scutt et al. 1990). De­ tection of SLR genes in both self-incompatible (Brassica and Raphanus) and self-compatible (Brassica, Arabidopsis) spp. and their high level of expression in both self-incompatible and self­ compatible plants indicate that SLR genes have no role in SI but may have a role in other post­ pollination processes. Recent studies have in­ dicated that SLR products may play a role in pollen adhesion (Luu et al 1999). Requirement of SLG and/or SRK in SI Re­ sponse: Initial analyses of some of the self-com­ patible mutants of B. campestris and B. oleracea showed drastic reduction in the level of stigma SLG products (Nasrallah et al. 1972, Nasrallah 1974) but not of SRK products. Other self-com­ patible mutants likewise showed SLG products comparable to SI lines but aberrant SRK tran­ scription (Nasrallah and Nasrallah 1993).These studies indicated involvement of both SLG and SRK genes in SI response in Brassica. Attempts to modify SI specificity in the stigma through transgenic expression of SLG/SRK alleles iso­ lated from plants of one S haplotype in plants of another S haplotype were not successful. High sequence similarity between SLG and SRK re­ sulted in homology-dependent gene silencing and reduced the expression of both transgene

and the endogenous gene pair; this resulted in self-fertile transgenic plants (Nasrallah 2000). It was thus not possible to attribute loss of SI to SLG or SRK. Subsequently two genomic clones that share SLG910and SRKgi0 were transformed into self­ compatible B. napus line Westar.The transgenic lines expressed both SLG910 and SRK910 at lev­ els comparable to that found in the wild line (Cui et al. 2000). The transgenic stigmas expressed the SI response in a haplotype specific manner but the pollen phenotype remained unchanged (see Brugiere et al. 2000). Recent transformation studies, however, have clearly shown that SRK but not SLG plays a criti­ cal role in SI responses (Takasaki et al. 2000). Either SRK28 or SLG28 was introduced into a homozygous S60 of B. campestris. Transgenic plants with only SRK28introduced rejected both S28and S^ pollen indicating the gain of S28speci­ ficity in the transgenic stigma; the pollen pheno­ type did not change. These experiments indi­ cated that SRK is the female S-specificity deter­ minant. However, further analysis indicated that SLG might enhance the SI recognition process. Plants expressing both SLG 28 and SRK 28 transgenes exhibited a stronger SI than those with only SRK,,,. It was suggested that SLG may play a role in stabilizing SRK or in enhancing diffusion of the pollen factor determining SI specificity (Brugiere et al. 2000, Franklin-Tong and Franklin 2000, Wheeler et al. 2001). Subsequent studies (Silva et al. 2001) have confirmed that SRK is the primary determinant of SI in the pistil. The SRK910and SLG910cDNAs isolated from SI B. napusYJ'\ line were individu­ ally transformed into SC B. napus. cv Westar. The SRK910transgenic Westar plants assumed the SI function and rejected W 1 pollen but SLGgi0 transgenic Westar plants were fully compatible to W1 pollen. Unlike earlier studies, transgenic plants expressing both SLG 910 and SRK910 showed no enhancement of SI reaction. Limited studies have been carried out to un­ derstand SI genes in Ipomoea trifida

152 Pollen Biology and Biotechnology

(Convolvulaceae), which also has SSI. Results have suggested the operation of an SI mecha­ nism in this species different from the one found in Brassica (see Wheeler et al. 2001).

Gametophytic systems S-allele specific glycoproteins from the styles which co-segregated with specific S-alleles have been identified in several gametophytic systems belonging to Solanaceae, namely Nicotiana, Lycopersicon, Petunia and Solanum (Bredemeijer and Blaas 1981, Clarke et al. 1985, Kheyr-Pour and Pernes 1985, Mau et al. 1986, Kamboj and Jackson 1986, Kirch et al. 1989, Xu et al. 1990a). All the S-proteins isolated to date from solanaceous members are glycosylated. They show considerable variabil­ ity in carbohydrate chains. The glycoproteins are basic (pi 7.5-9.5) and their molecular weights range from 27-33 kD.The putative S-allele spe­ cific proteins accumulate late in pistil develop­ ment. All the S-allele specific proteins have 15 highly conserved N-terminal amino acids (Mau etal. 1986). Most of the initial studies on cloning of Sgenes were carried out by Clarke and her asso­ ciates on N. alata (Anderson et al. 1986,1989, Cornish et al. 1987, Dodds etal. 1993). AcDNA clone encoding a stylar glycoprotein of N. alata which co-segregated with S2-allele was the first to be isolated (Anderson et al. 1986). Subse­ quently, cDNAs for several other S-glycoproteins of N. alata were cloned. 32P-labelled cDNA en­ coding S2 glycoproteins was used to detect S2 mRNA by in-situ hybridization using sections of the pistil; expression of the S-gene was shown to occur in the stigma and throughout the trans­ mitting tissue. Similar localization of S2specificglycoproteins was also shown through immunocytochemistry using antibody raised against specific S-glycoprotein (Cornish et al. 1987, Anderson etal. 1989). cDNA clones corresponding to several S-alleles of P. hybrida (Clark et al. 1990) and P. inflata (Ai et al. 1990) and two of Solanum chacoense

(Xu et al. 1990b) have also been obtained. Amino acid sequence alignments of different S-glyco­ proteins have shown five highly conserved re­ gions and several hypervariable regions. S-locus is highly polymorphic and shows a high de­ gree of diversity between S-allele glycoproteins in hypervariable regions (loerger et al. 1990, Kheyr-Pour et al. 1990). For example, S,, S2, S3 and S6 glycoproteins from N. alata share only 51.5% overall sequence identity (Haring et al. 1990, Clark and Kao 1991). One of the important findings with respect to the solanaceous S-gene is that the amino acid sequences of the S-gene product in two of the conserved regions are homologous to the re­ gions conserved in the ribonucleases (RNases) (McClure et al. 1989); they are close to the cata­ lytic domain of RNases and include histidine residues which are critical for enzyme activity. Further studies showed that all S-glycoproteins from solanaceous species are functional RNases (Clark et al. 1990, Singh et al. 1991, Xu et al. 1990b) and are referred to as S-RNase proteins. Later studies revealed that S-RNase entered the pollen tubes and degraded r-RNA in incom­ patible tubes but not in compatible tubes indi­ cating that S-RNase functioned enzymatically in the SI reaction (McClure et al. 1990). How­ ever, there was no S-allele specificity in the ac­ tivity of S-RNase in vitro grown pollen tubes (Gray et al. 1991). S-gene products in Pyrus, another gametophytic system, belonging to Rosaceae, have also been shown to be RNases (Sassa et al. 1992, Certal et al. 1999). Subse­ quent studies have shown that in members of Scrophulariaceae also S-gene products of the pistil are RNases (Xu et al. 1996) Studies on genetic transformation in Petunia and Nicotiana have provided direct evidence for the involvement of putative S-genes in SI re­ sponse. Transformation of S,S2 plant with an S3 gene resulted in transgenic plants which rejected S3 pollen; such plants had S3 proteins in addi­ tion to S, and S2 proteins at comparable levels.

Self-incompatibility 153

Other transgenic plants which displayed partial seed-set with pollen had smaller amounts of S3 proteins (Lee et al. 1994). Ribonuclease activity is essential for rejec­ tion of self-pollen (Huang et al. 1994). S3 gene of P. inflata was mutated by replacing the codon for His-93 (which is essential for ribonuclease activity) with a codon for asparagine, and the mutant gene was transferred to the plants of S,S2 genotype. Two transgenic plants produced a mutant S3 protein at a level comparable to that of S3 protein in SI plants, but did not exhibit RNase activity. These two transgenic plants failed to reject S3 pollen. In Nicotiana also transformed plants with SA2 allele showed high levels of SA2 RNase and re­ jected SA 2 pollen (Murfett et al. 1994). In Lycopersicon peruvianum (Royo et al. 1994), loss of histidine residue at the active site of Slocus ribonuclease has been shown to be asso­ ciated with self-compatibility.

Papaver system Papaver rhoeas, a gametophytic SI system, shows distinct differences from other gameto­ phytic systems. Inhibition takes place on the stigma, similar to sporophytic systems. Using in vitro assay, stigmatic S-gene products have been isolated and characterized (Franklin-Tong et al. 1989, Franklin-Tong and Franklin 1992).The pro­ teins which co-segregate with the S alleles have been purified (Foote et al. 1994). The putative S1 products are small extracellular molecules (ca. 15 kD, Foote et al. 1994). These S, prod­ ucts also correlated developmentally with the expression of SI. Unlike other gametophytic sys­ tems in which S-glycoproteins form a major com­ ponent of pistil proteins, in P. rhoeas they form only ca 0.5-1.0% of the total proteins of stig­ matic papillae. More importantly the S-gene products in P. rhoeas are not ribonucleases, in­ dicating that the mechanism of SI operation is different in this system from other gametophytic systems (Franklin-Tong et al. 1991). Several alleles of P. rhoeas and P. nudicaule have been cloned (Foote et al. 1994). The se­

quence predicts the position of a single poten­ tial glycosylation site. Expression of S allele is confined to the stigmatic tissue and is largely in the mature stage.

S-allele-specific components in pollen In contrast to the progress made in understand­ ing the details of S-alleles and their products in the pistil, our knowledge of S-allele components in the pollen is very meagre. Attempts to identify pollen proteins which co-segregate with S-alle­ les have not been successful. In Brassica, a low level of expression (several hundred fold lower than in stigmas) of SLG product was detected in post-meiotic anthers but not in mature pollen (Nasrallah and Nasrallah 1989, Guilluy et al. 1991). Also, transgene-induced silencing of SLG (Toriyama et al. 1991) or SLG and SRK (Conner et al. 1997, Stahl et al. 1998) that resulted in the breakdown of SI in the stigma, did not affect the pollen phenotype, indicating that SLG and SRK do not function as S-genes in the pollen. Doughty et al. (1998) reported a 7 kD pollen coat protein that forms a complex with stigmatic extracts. However, further studies indicated that this in­ teraction was not S-allele specific. Studies on Nicotiana alata (Dodds et al. 1993) have shown the expression of a transcript ho­ mologous to the cDNA of the S2 gene in devel­ oping pollen. The level of expression was only ca 1% that found in the style. Evidence has also been obtained for the presence of S-proteins in the intine of mature pollen. None of these stud­ ies on Nicotiana, however, have categorically demonstrated the involvement of any of these pollen components in SI. Recently a pollen S-gene from Brassica oleracea was identified and designated as SCR (S-locus cysteine-rich protein) (Schopfer et al. 1999). SCR is expressed exclusively in anthers during pollen development and encodes a short open reading frame with an unusually high fre­ quency of cysteine residues; it accumulates af­ ter meiosis. SCR product is a hydrophilic pro­ tein of 8.4-8.6 kD. So far over 20 SCR alleles have been identified.Takayama et al. (2000) also

154 Pollen Biology and Biotechnology

isolated an analogous gene in B. campestris and designated it as SP 11 (S-locus protein 11 ). The following data have confirmed that SCR is indeed the pollen S-gene: (i) A self-compatible mutant of B. oleracea in which SI breakdown was confined to the male determinant did not have a detectable amount of SCR expression, (ii) Deduced amino acid se­ quence analyses of SCR proteins from S6, S8, and S 13 haplotypes showed a high degree of polymorphism giving an overall similarity of only 30-42%. (iii) Transformation studies involving transfer of SCR6coding region to S2S2 homozy­ gous B. oleracea have also confirmed the role of SCR in SI. The pollen from S2S2/SCR6+ transformants acquired S6specificity. SCR is expressed gametophytically in mi­ crospores. It is suggested that the proteins of the two SCR alleles are synthesized during pol­ len development and are incorporated into the pollen coat as a mixture when pollen grains get mixed within the anther locule. In-situ hybridiza­ tion studies have indicated that SCR is ex­ pressed in the tapetal cells also, suggesting that this gene is expressed both sporophytically and gametophytically (Takayama et al. 2000). Appar­ ently, the pollen coat substances, derived from the tapetum, would contain the products of both the S-alleles (see Nasrallah 2000).

Basis of S-allele Specificity In spite of the progress made in understanding the details of S-genes in Brassica and members of Solanaceae, the basis of S-allele specificity is not clear. In Brassica, SLG protein has two highly conserved and two highly polymorphic regions. In Nicotiana, there are five conserved regions which are important for S-RNase activ­ ity and several hypervariable regions. The vari­ able regions are likely to be involved in allelic specificity. Although S-allele proteins of both Brassica and Nicotiana are glycosylated, there is no evidence to suggest that carbohydrate chains have any role in S-allele specificity (Broothaerts et al. 1991, Karunanandaa et al. 1994).

Mechanism of Pollen Recognition and Inhibition There is no complete understanding either on the basis of recognition or on the mechanism of inhibition in any SI species. However, on the basis of available evidence, three different mod­ els have been proposed: (i) Brassica and other sporophytic systems, (ii) solanaceous gameto­ phytic systems, and (iii) members of Papaveraceae. A summary of the three models suggested (Dickinson 1994, Wheeler et al. 2001, Stone and Goring 2001) follows. In Brassica studies described earlier indi­ cated that SI response requires a functional in­ teraction between SLG and SRK products. Ac­ cording to Nasrallah and Nasrallah (1993), there are two possibilities for such an interaction: (i) the SLG may become competent for binding with SRK only after it is modified by pollen-borne Sallele product. In this model, pollen S-product acts as a modifying factor for SLG and (ii) the Sallele product acts as an extracellular ligand to bind SLG and SRK.The binding of SLG and SRK activates the receptor kinase in the stigma pa­ pillae which then initiates phosphorylation of one or more intermediates that ultimately produce a localized response within the stigmatic papillae that inhibits the pollen.This model envisages the primary site of inhibition reaction to be the pa­ pillae, unlike in gametophytic systems in which the pollen seems to be the site of inhibition re­ action. Also, the model explains neither the ba­ sis of S-allele specificity nor the mechanism of inhibition. Recent evidence indicating that only SRK is involved in SI response in the pistil and SCR is the pollen S-gene has led to reinterpretation of SI inhibition in Brassica (Brugiere et al. 2000, Franklin-Tong and Franklin 2000). Following pol­ lination, pollen coat substances which include SCR products (and many others involved in other aspects of post-pollination events such as pol­ len adhesion) flow from the pollen, form a me­ niscus and come into contact with SLG and/or SRK products. The SCR and extracellular

Self-incompatibility 155

domain of SRK protein encoded by the same haplotype interact and establish recognition (Fig. 9.10). This interaction is likely to result in activa­ tion of the intracellular serine-threonine protein kinase domain of SRK. Activated kinase domain phosphorylates ARC 1 (arm-repeat containing protein 1 , a stigma specific molecule that is phosphorylated by SRK in vitro), which in turn ini­ tiates an intracellular signalling cascade in the papillar cell and eventually inhibits pollen ger­ mination or pollen tube growth. ARC 1 has been shown to be required for the SI response in Bras-

sica. Transgenic plants of B. napus in which ARC 1 expression was down-regulated by an antisense approach showed reduced ability to reject self-pollen (Stone et al. 1999). On the ba­ sis of indirect evidence it has been suggested that signal transduction pathway may finally regulate the activity of aquaporins, water chan­ nel proteins, in the stigmatic papilla to limit the availability of water to self-pollen. As pointed out earlier, the role of SLG protein in SI response is not yet clear; it may play a role in stabilizing the SRK or in enhancing diffusion of pollen SRC

pollen grain Si j s3 exine

S-| pollen product (SCR) S3 pollen product (SCR) St SLG meniscus


S2 SLG s2 s rk

Stigma papilla Si

phoshorylation cascade

cell wall

plasma membrane

Fig. 9.10 Proposed model for operation of SI in sporophytic system ( Brassica). In pollen grain, the products of two S-alleles, S, (matching allele with that of the stigma) and S3 (non-matching allele with that of the stigma) are shown. S,-pollen product (SCR) binds to the St-SRK (and possibly SLG) product on the stigmatic papilla. This recognition leads to phosphorylation cascade that eventually inhibits pollen germination. S3-SCR does not match with the S-allele products of the stigma and thus does not activate phosphorylation cascade.

156 Pollen Biology and Biotechnology

the basis of specificity. It is suggested that ei­ ther RNase enter only the self-pollen tubes as a result of recognition and not the cross-pollen tubes, or the cross-pollen tubes somehow inac­ tivate RNase. Recent immunological studies (Luu et al. 2000) indicate that uptake of S-RNase is inde­ pendent of S-genotype. S12 pollen was used to pollinate pistils of S^S^ genotype. S „ antibod­ ies detected markedly higher amounts of S1( RNase within the pollen tubes compared to the extracellular matrix of the pistil. The evidence in Papaver rhoeas, in which Sallele products are not RNases, suggest that the operation of SI in this species differs compared to other gametophytic systems (Franklin et al.

products or may act as accessory receptor or have a more general role in pollen-pistil interac­ tion. The model suggested for members of Solanaceae, Rosaceae and Sterculiaceae, is based on the fact that S-allele products in the pistil are functional RNases. According to this model, the S-RNase of the transmitting tissue enters the self-pollen tubes (Fig. 9.11). The internalized SRNase then degrades the rRNA in the pollen tube, leading to cessation of protein synthesis and consequent arrest of pollen tube growth (Newbigin etal. 1993, Dickinson 1994). Although all the data including transformation experiments support the model on the mechanism of pollen tube inhibition, there is no evidence to explain

5 1 RNase Pollen tube S 1 : S3 cell of transmitting tissue

5 2 RNase S 1 RNase receptor 5 3 RNase receptor

RNA / ^degradation

Fig. 9.11 Proposed model for SI in solanaceous species such as Nicotiana and Petunia in which S-allele product in the pistil is RNase. Receptors of both S, and S3alleles are depicted in the pollen tube. S,-RNase released into the extracellular matrix of the pistil is recognized and internalized by the matching S,-receptors on the pollen membrane; S,-RNase degrade the rRNA in the pollen tube leading to cessation of protein synthesis and arrest of pollen tube growth. According to the model depicted, when S-RNase (S2) of the ECM does not match with the S-allele receptor (S3) of the pollen, the S-RNase fails to enter the pollen tube.

Self-incompatibility 157

1995). According to the model for Papaver (Franklin-Tong et al. 1993, Dickinson 1994, Franklin et al. 1995), stigmatic S-protein acts as a signal molecule and binds with homologous S-receptors of the pollen, believed to be located in the plasma membrane (Fig. 9.12). Inhibition of self-pollen is rapid and is brought about within minutes. The signal molecule (S-allele product of the stigma) is not internalized but the binding triggers a calcium-mediated signal transduction pathway inside the pollen, leading to the release of Ca2+from internal stores within a few seconds

of SI interaction. This transient increase of intra­ cellular Ca2+is the initial step and leads to inhibi­ tion of pollen tube growth. Evidence for the operation of Ca2+-mediated signalling during SI response in P. rhoeas has come from studies using Ca2+-selective dye, Calcium Green-1 , which has high affinity and selectivity for free calcium, in the in vitro assay (Franklin-Tong et al. 1993). Calcium Green un­ dergoes a marked change in fluorescence on binding to cytoplasmic free Ca2+. Pollen tubes growing in vitro were microinjected with the dye

pollen grain/tube S* . So

transient rise V in Ca2+ ♦

product stigma sti

S2 product


Sf receptor


S3 receptor



♦ 9i • 9I

.V .


ú \ w

I *





stigma papilla #

Si s2

Fig. 9.12 Proposed model for operation of SI in Papaver system. Both S, and S3 receptors are depicted in the pollen. Binding of S^allele product of the stigma to the S^allele receptor present on the pollen membrane results in the activation of calcium­ mediated signal transduction pathway leading to transient release of calcium from internal store, increased levels of intracel­ lular calcium inhibit pollen germination/pollen tube growth.


Pollen Biology and Biotechnology

and cytosolic free Ca2+was monitored using la­ ser scanning confocal microscopy. Addition of S-glycoproteins from self-stigma in the medium induced a transient increase in the level of cyto­ solic free Ca2+ in the pollen tubes and resulted in the arrest of their growth. No rise in Ca2+ lev­ els was detected after addition of protein from compatible pistil or heat denatured S-glycoproteins.The photoactivation of bound Ca2+to arti­ ficially elevate cytosolic free Ca2+ also resulted in the inhibition of pollen tube growth, similar to SI response. These studies clearly establish a link between transient rise in cytosolic free cal­ cium and pollen tube inhibition and the media­ tion of SI inhibition by Ca2+. Calcium is impor­ tant for the regulation of pollen tube growth (see Chapter 4) and is known to act as a second mes­ senger in plant cells. Further studies have indicated that the initial Ca2+signal results in phosphorylation of several proteins, some Ca2+-dependent, and others Ca2+-independent (Rudd et al. 1997). SI re­ sponse at a later stage seems to target the ac­ tin cytoskeleton of pollen tubes (Geitmann et al 2000). In actively growing pollen tubes, F-actin bundles are generally absent at the tip region; within 1-2 min of SI reaction, there is a reorga­ nization of F-actin bundles at the tip of self-pol­ len tubes. In another 8-10 min, fine fragments of actin are seen throughout the cytoplasm indi­ cating its disorganization. Nuclear fragmentation, a feature of programmed cell death, has been detected 4-12 h after SI response (Wheeler et al. 2001).

HETEROMORPHIC SI In heteromorphic SI, plants of the same spe­ cies produce more than one type of flower. The most prominent difference between floral mor­ phs is the relative position of the anthers and the stigma. Floral morphs may be of two (dimorphic/distylous) or three (trimorphic/tristylous) types in a given species but each plant produces only one type of flowers. According to de Nettancourt (2001) heteromorphic incompati­

bility is recorded in 22 families of flowering plants. Evolution of heterostyly has been discussed by many investigators. According to Baker (1966), Charlesworth and Charlesworth (1979) and Ganders (1979), dillelic SI evolved first and the morphological differences were superim­ posed on pre-existing SI. According to Lloyd and Webb (1986) and Barrett (1990), however, mor­ phological differences preceded physiological SI, and the latter evolved by gradual adjustment of pollen tube growth to different morphs. On the basis of its taxonomic distribution as well as the prevailing diversity of structural and physiologi­ cal mechanisms, most of the investigators sup­ port polyphyletic origin of heterostyly.

Dimorphic Systems The two floral forms of dimorphic species are referred to as thrum/short-styled and pin/longstyled morphs. The anthers in one morph corre­ spond to the level of the stigma in the other morph. Pollination between flowers of different morphs is compatible (intermorph compatibility) while those within the morph are incompatible (intramorph incompatibility). Intramorph incom­ patibility and intermorph compatibility transgress species limits. In Linumgrandiflorum, intramorph inter-specific pollinations are invariably incom­ patible while intermorph pollinations are com­ patible (Ghosh and Shivanna 1984b).The floral morphs of the two types and their compatibility relationships are diagrammatically presented in Figure 9.13A. Dimorphic condition is common in members of Rubiaceae, Plumbaginaceae, Linaceae and Boraginaceae (de Nettancourt 1977, Ganders 1979). Distyly is controlled by a single gene com­ plex, S, with two alleles S and s. The allele for short-style (S) is dominant over that of long-style (s). Long-styled morphs are homozygous reces­ sive (ss) while short-styled morphs are heterozy­ gous (Ss).Thus, intermorph compatible pollina­ tions (ss x Ss) produce progeny of approximately equal number of short- and long-styled morphs. Pollen with recessive s allele from a short-styled

Self-incompatibility 159 incompatible


namentation of the pollen and morphology of the stigmatic papillae (Figs. 9.14 and 9.15, Table 9.2). Ail these differences may not be present in all the dimorphic species. Table 9.2 Morphological differences between thrum and pin morphs in some dimorphic species (based on data presented in Dulberger 1992; see Dulberger 1992 for details of refer­ ences) Character



Exine sculpture

Pollen production (No. of pollen grains per flower) Anther size

Pollen colour






B Fig. 9.13 Diagram of heteromorphic incompatibility. A. Di­ morphic system. B. Trimorphic system. Compatible pollina­ tions are depicted by solid arrows and incompatible pollina­ tions by dashed arrows (after Shivanna 1982).

morph is compatible on a long-styled morph (ss) but incompatible on a short-styled one (Ss). Thus, incom patibility in the pollen is sporophytically determined. The style length ratios of pin and thrum mor­ phs are variable; for example, 4:1 in Primula auricula and 2:1 in Limonium vulgare. Differ­ ences in the length of the style and stamens in the two morphs are associated with other mor­ phological differences such as size and/or or­

Some examples

Relative lengths of Longer style and All dimorphic style and stigma shorter stamen in species pin, shorter style and longer stamen in thrum Pollen size



Stigma surface

Stigmatic papillae

Wall of papillae

Thrum pollen larger than pin pollen

Armería marítima Amsinckia spp. Oldenlandia spp. Primula spp. Pin pollen larger Linum spp. than thrum pollen Limonium spp. Variable in different Armería marítima Limonium vulgare species Linum spp. More in pin morph Amsinckia grandiflora Primula spp. Linum spp. Thrum anthers larger Pin anthers larger Amsinckia spp. Thrum pollen dark Linum blue, pin pollen grandiflorum dark grey Linum tenuifolium Pin pollen pale brick-red and thrum pollen yellow Thrum stigma wet, Linum pin stigma dry grandiflorum Pin stigma wet, Primula spp. thrum stigma dry Pin stigma nonLimonium sp. papillate, thrum stigma papillate Linum spp., Papillae of pin morph longer/ Primula spp. larger Linum pubescens Thrum papillae with subcuticular pectin cap, absent in pin papillae

160 Pollen Biology and Biotechnology


B Pin-morph

Fig. 9.14 Differences in structural features of pollen and stigma in a dimorphic SI system, Limonium vulgare. A-C. Thrum morph (A) with non-reticulate pollen exine (B) and papillate stigma (C). D-F. Pin morph (D) with reticulate pollen exine (E) and non-papillate stigma (F) (after Baker 1966).

The S-gene in heteromorphic species has been traditionally considered a supergene made up of several closely linked genes. According to

Lewis (1954) the S-gene complex comprises six linked genes: G, S, I', I", P and A (G = style length, S = stigm a surface, I' = pollen

Self-incompatibility 161

Fig. 9.15 Stigmatic papilla of pin and thrum morphs in Linum grandiflorum. A, B. Cytochemical localization of non-specific esterases on the stigma surface of pin (A) and thrum (B) papillae. In A the pellicle (arrowheads) is smooth and continuous while in B, it is raised and ruptured at places because of the accumulation of secretion product just below the pellicle (aster­ isk). C, D. Scanning electron micrographs of stigmatic papillae of pin (C) and thrum (D) stigmas. The pin stigma is of the dry type and that of the thrum stigma, the wet type. Scanning electron micrographs of pollen grains of thrum (E) and pin (F) morphs of Linum austriacum. The exine processes are large and uniform in thrum pollen while they are of two sizes in pin pollen. Insets show magnified portions of exine of respective pollen grains (after Ghosh 1981).

162 Pollen Biology and Biotechnology

incompatibility, \" = stylar incompatibility, P = pol­ len size and A = stamen height). Incompatibility genes are closely linked to genes determining floral polymorphism and thus are inherited to­ gether. Several studies have indicated physi­ ological and biochemical differences between

the pollen of pin and thrum morphs (Fig. 9.16, Table 9.3). Compared to the number of studies on the structural differences between thrum and pin morphs, relatively few studies have been carried out on physiological and biochemical aspects. In the light of these physiological differ-

100 I



0 CD

1 80 a >) O

so 0 s




0 1 o
S c-> S c’ -> SC. They explained the growth of pollen tubes of SC species in the pistil of SI species reported in some crosses on the basis that the pollen has an Sc allele and has not yet stabilized to an SC allele. Pandey (1968, 1969, 1970) elaborated the dual function hypothesis further to accommo­ date all Ul irrespective of the presence of SI in one of the parents. According to him, the S-gene complex has two specificities—primary speci­ ficity and secondary specificity; the former con­ trols interspecific incompatibility and the latter, SI. S-gene-mediated responses of Ul are often modified by other genes (Pandey 1968, de Nettancourt 1977). According to the second hypothesis, SI genes play no role in Ul (Grun and Radlow 1961, Grun and Aubertin 1966, Abdalla 1974). Inhibition of the pollen carrying the SC allele is mediated by specific genes different from the S-locus. Accord­ ing to Abdalla (1974), there are specific Ul genes that inhibit fertilization by pollen grains carrying specific SC alleles. In the absence of matching Ul genes, the cross between SI x SC plants will be successful in both directions. In the presence of matching Ul genes, the cross is successful in one direction only (SC x SI). Crosses between distantly related species will fail in both direc­ tions, irrespective of the nature of incompatibil­ ity (SC x SI, SI x SI, SC x SC) because of other barriers. There are hardly any studies on the physi­ ological aspects of Ul. A few studies carried out on the cytological details of pollen tube inhibi­

tion have shown that the inhibition of pollen in Ul is stronger than that in SI (Nicotiana—Pandey 1964, Solarium—Grun and Aubertin 1966; Lycopersicon— Lewis and Crowe 1958, Hogenboom 1972). In the pistil of N. alata, for example, selfed pollen tubes are inhibited after growing through 10-27% of the style length, whereas pollen tubes of N. longsdorfii, a self­ compatible species, are inhibited in the stigma (of N. alata) before entering the style (Pandey 1964).

Incongruity/Passive Inhibition As discussed earlier (Chapter 9), SI inhibition is the result of active recognition. Self-pollen is actively recognized as a result of interaction of S-allele products; positive recognition results in activation of metabolic processes in the pollen and/or the pistil and brings about pollen inhibi­ tion. Compatible pollination activates many physiological processes in the pistil, as a result of co-adaptation between the pollen and the pis­ til, which facilitates sequential completion of all post-pollination events. In most interspecific crosses, however, the arrest of post-pollination events seems to be passive as a result of lack of co-adaptation be­ tween the pollen and the pistil. Such an incom­ patibility which is passive (and not the result of active recognition) has been termed incongru­ ity (Hogenboom 1975). Compatibility requires matching of the genetic system between the pistil and pollen; for each barrier gene or gene com­ plex active in the pistil, there must be a penetra­ tion gene or gene complex in the pollen (Table 10.2). Incongruity is due to the absence of ge­ netic information in one partner for some rel­ evant character of the other and the details vary in different crosses. Incongruity is a by-product of evolutionary divergence. Among a freely interbreeding popu­ lation, a subpopulation may differentiate as a result of changed environment. This may bring in an extra barrier in the pistil of one of the popu­ lation for the pollen of the other, resulting in in­ congruity between the two populations. The

Interspecific Incompatibility 169 Table 10.2 Compatibility relationships based on incongruity concepts of Hogenboom (1975). Presence/absence of pen­ etration capacity of pollen to the barriers present in the pistil determines whether the cross is compatible (+) or incom­ patible (-) Barriers in pistil


Penetration capacity of the pollen for A




+ -

+ +





+ + + -

+ + + +

barriers increase as the divergence of the two populations increases. The incongruity operates not only during pollen-pistil interaction, but also during post-fertilization stages. Thus, incongru­ ity is comparable to the lock and key mecha­ nism; absence of a suitable key(s) with the pol­ len for the lock(s) present in the pistil results in incongruity. Incongruity is a complex phenom­ enon and may involve many genes depending on the extent of genetic divergence. The greater the divergence, the earlier the barrier. The hypotheses put forward to explain inter­ specific incompatibility are based mostly on ge­ netic studies. There is very little information on physiological and biochemical aspects. In the cross Petunia hybrida x Salpiglossis sinuata, the morphological abnormalities of incompatible pollen tubes (branching, swelling of tube tip, in­ creased callóse deposition etc.) and the pattern of respiration in pollinated pistils were similar to those following SI pollinations (Roggen and Linskens 1967). On the basis of these results, Roggen and Linskens (1967) suggested that the mechanism of pollen tube inhibition is similar in both SI and interspecific incompatibility. In Lilium longiflorum, however, unlike SI which can be overcome by giving hot water treatment to the pistil or exposing the plants to elevated tempera­ tures, interspecific incompatibility could not be overcome by either treatment (Ashcer and Peloquin 1967, Ascher 1975,1979). The evidence available so far suggests that interspecific incompatibility between closely re­

lated species involving one or both SI parents; pre-fertilization inhibition is likely to be mediated through S-genes. This is well illustrated in heteromorphic species of Linum. Studies conducted by Ghosh and Shivanna (1984b) and Murray (1986) on different species of Linum have shown that intermorph pollinations (pin x thrum, thrum x pin), both within as well as between species, are compatible whereas intramorph (pin x pin, thrum x thrum) pollinations are incompatible. Thus, intermorph compatibility and intramorph incompatibility operate within as well as between species. Obviously pollen tube inhibition in such interspecific crosses is mediated through the S­ gene complex. Subsequent post-fertilization barriers, such as embryo and/or endosperm breakdown are, apparently, controlled by other genes. In distantly related species, SI genes are not involved in incongruity even when one or both the parents are self-incompatible. In such crosses all individuals of a species regardless of the S-genotype of the plants behave similarly for incongruity. Incongruity operates even when SC accessions of SI species are used in the crosses. Many recent studies on wide hybrid­ ization, particularly on Brassica, have shown that Ul operates more often in crosses between the cultivars (female) and wild species (male) than in the reciprocal combination, irrespective of the presence or absence of SI in one of the species (see Shivanna 2000). Such Ul in which S-genes do not have a role has been termed unilateral incongruity (Liedl et al. 1996). According to Sampson (1962), successful pollen-pistil interaction depends on the degree of complementation between stigmatic and pol­ len molecules. There are two areas—‘species area’ and ‘S-allele area’—in which complemen­ tation can occur. Complementation at the S-al­ lele area results in SI due to S-allele identity. In­ terspecific crosses are compatible only when there is complementation at the species area, but not at the S-allele area. Thus complementa­ tion at both areas or none of the areas results in incompatibility (see also Heslop-Harrison 1975).

170 Pollen Biology and Biotechnology

De Nettancourt and colleagues (1973a, b, 1974, 1975) studied ultrastructural details of pollen tube inhibition in SI ( Lycopersicon peruvianum) and UI (L. peruvianum x L. esculentum). Pollen tubes in both types of polli­ nation accumulated a large number of bipartite particles at the tip and formed characteristic concentric endoplasmic reticulum. Consistent differences were observed between the pollen tubes inhibited in SI and those in Ul only in de­ tails of the outer wall; it was thick in the former but rather thin in the latter. Based on these and other genetic studies de Nettancourt (1977) sug­ gested that SI and Ul are two distinct but closely related rejection processes, and Ul results from the interaction of S-elements in the pollen grains with products of one or several unidentified stylar genes.

Mechanism of Passive Inhibition Figure 10.1 summarizes different levels at which passive inhibition may operate during pollen-pis­ til interaction and possible mechanisms. As pointed out earlier (Chapter 8), all post-pollina­ tion steps require co-adaptation between the pollen and the pistil for successful completion of pollen-pistil interaction. Lack of such a co­ adaptation at any stage of pollen-pistil interac­ tion would result in inability of the pollen to com­ plete post-pollination events (Fig 10.2A-C). Pollen adhesion and hydration depend on morphological and physiological co-adaptation between the pollen and the stigma. The amount and nature of secretion products on the stigma and pollen coat play an important role in pollen adhesion. Absence or insufficient amount of these components in one of the partners will result in failure of pollen adhesion. For example, Arabidopsis thaliana stigma selectively binds A. thaliana pollen with higher affinity than pollen from related species (Zinkl et al. 1999). Pollen grains from other dicotyledonous species could be washed off by detergent treatment of the stigma; such treatment has little effect on the binding capacity of the pollen of the same spe­ cies (Zinkl and Preuss 2000). This is obviously

due to lack of specific components on the stigma/ pollen required for adhesion of pollen of other taxa. In the absence of effective pollen adhe­ sion, pollen hydration and, consequently, pollen germination would be prevented. Pollen hydration is the result of water uptake from the stigma mediated through differences in the osmotic potential of the pollen and the stigma. Mismatch in the optimal osmotic poten­ tial between pollen and stigma results in lack of hydration or insufficient hydration. Lack of co­ adaptation between the supply and demand position would also result in insufficient hydra­ tion. This would happen particularly in crosses involving the male parent with large pollen and the female parent with small pollen; the stigmatic papillae may not be able to provide sufficient water for full hydration of larger pollen. Alterna­ tively the limited number of germ pores of the large pollen may not establish effective contact with the stigmatic papillae (Heslop-Harrison 1979d). Recent studies have given a clear indi­ cation that some of the components of the pol­ len coat are required for pollen adhesion (see Chapter 8). Failure of pollen to germinate on the stigma or the entry of pollen tubes into the stigma is one of the most common barriers observed, particularly in wide crosses (Martin 1970, Knox et al. 1976, Stettler et al. 1980, Batra et al. 1990, Barone et al. 1992, Gundimeda et al. 1992).The causative factors have not been worked out in most of the crosses. However, many studies on pollen-pistil interaction have indicated some of the possible causes for lack of pollen germina­ tion and pollen tube entry. As pointed out earlier, lack of pollen adhe­ sion or hydration would result in the failure of pollen germination. Even in those crosses in which adhesion and hydration are normal, pol­ len grains require suitable conditions/factors for germination. Pollen grains of many species are known to require boron and/or calcium for ger­ mination and the stigma provides these compo­ nents to the pollen (Bednarska 1991). In Vitis vinifera boron-deficient stigmas fail to support

Interspecific Incompatibility 171 Pollen lands on the stigma Lack of morphological complementation Absence of specific adhesion components on pollen/stigma Pollen adhesion

Differences in osmotic potential of pollen and stigma Lack of co-adaptation in supply and demand position

Pollen hydration

Absence of specific components required for hydration on pollen /stigma Absence of pollen germination factors on stigma Insufficient/ uncontrolled hydration of pollen

Pollen germination Absence of activators required for the precursors of pollen cutinases Lack of suitable lipids on stigma/pollen required to provide directional supply of water to pollen Pollen tube entry into the stigma (in dry stigma)

Inability of pollen tubes to utilize stylar nutrients Absence of specific TTS proteins required for pollen tube growth

>r Pollen tube growth through the style

Differences in the length of styles of the two parents

Lack of guidance cues from the synergids to attract pollen tube yr

Pollen tube entry into ovules Fig. 10.1 Stages in passive rejection and probable mechanisms (modified from Shivanna and Johri 1985).

pollen germination (see Chapter 4). Thus, lack of the required amount of boron and/or calcium on the stigma may be a cause for failure of pol­ len germination. Suitable pH is also an impor­ tant requirement for pollen germination (Rob­ erts et al. 1983, Ganeshaiah and Shaankar 1988). Stigmatic exudates of different species

often show marked variation in pH. Species of Rosa, for example, show a pH variation of 5 to 9 (Gudin and Arena 1991). Such pH differences may act as important barriers. Many flavonols are required for pollen ger­ mination and pollen tube growth (see Chapter 8). This requirement is met either by the pollen

172 Pollen Biology and Biotechnology

Fig. 10.2 Fluorescence micrographs of pistils stained with decolourized aniline blue following intergeneric crosses. A, B. Brassica campestris x Diplotaxis berthautii. Pollen grains have germinated on the stigma (sg) but pollen tubes have failed to enter the papilla. B is a magnified view of one of the pollen grains that has induced the development of callose in the stigmatic papilla (arrowhead). C. Brassica carinata x Sinapis puoescence. Some of the pollen tubes have entered the papillae (arrow­ head) but are inhibited soon after. D-F. Erucastrum abyssinicum x B. juncea. Pollen germination (D), pollen tube growth in style (E) and pollen tube (arrow head) entry into ovule (F) are normal. Thus this cross has no prefertilization barriers (A, B— after Vyas 1993; C—after Ranbir Singh, unpublished, D-F—after Rao 1995).

Interspecific Incompatibility 173

or the stigma. Flavonol requirement would not be met in a cross in which pollen and stigma are both deficient in flavonols. Some components of the stigma surface proteins/glycoproteins are also required for pollen germination in some species (Heslop-Harrison and Heslop-Harrison 1975). Lack of such proteins would inhibit pol­ len germination. Many studies have shown that phenolic com­ pounds of the stigma may play an important role in passive inhibition; they selectively inhibit or promote germination (Martin 1969,1970,1972, Martin and Ruberte 1972, Sedgley 1975, Tara and Namboodiri 1976). Of the three male sterile mutants of Impatiens sultani, orange, crimson and pink, only orange and crimson set seeds upon pollination, but not the pink (Tara and Namboodiri 1976).This was due to the failure of the pink stigma to support pollen germination. Chromatographic studies showed that the two phenolic spots present in the stigmas of orange and crimson were absent in the pink (Tara and Namboodiri 1976, Bhaskar and Namboodiri 1978). The authors suggested that absence of these phenoiics in the stigma of the pink mutant was responsible for its failure to support pollen germination. Uncontrolled hydration, as happens on a wet stigma with aqueous exudate, may be another cause for failure of pollen germination. Several in-vitro germination studies have shown the im­ portance of controlled hydration for successful germination of pollen (see Chapter 4). This is particularly true for those species characterized by dry stigma (Heslop-Harrison and Shivanna 1977, Shivanna and Heslop-Harrison 1981). Pollen germination will fail in crosses involving pollen requiring controlled hydration and the stigma lacking adaptation for controlled hydra­ tion. Recent studies have shown that pollen ad­ hesion, hydration and germination are more complex and a number of components on the surface of both pollen and the pistil play critical roles in these processes (see Chapter 8). Of these the proteins and lipids present on the

pollen as part of the pollen coat substances and/ or the stigma surface are required for these pro­ cesses, especially in species with dry-type stigma. Several mutants defective in one of these components have been isolated in Arabidopsis (see Chapter 8); these mutations act as barri­ ers to pollen function. Entry of the pollen tube into the stigma is criti­ cal. In species with a dry stigma covered with a layer of cuticle, pollen tubes enter the latter by activation of pollen cutinases. Pollen cutinases seem to require activators of the stigma surface. In some species, treatment of the stigma with proteases (Heslop-Harrison and HeslopHarrison 1975) or concanavalin A (Knox et al. 1976) does not affect pollen germination but to­ tally inhibits pollen tube entry, indicating that some components of the pellicle are required for effective operation of the pollen cutinases. Results of recent studies have demonstrated that several species with wet as well as dry type stigma require long-chain lipids, which facilitate directional supply of water to the pollen and thus guide the resultant pollen tubes into the stigma (see Chapter 8). In wet stigma such as Nicotiana, these lipids are present in the stigmatic exu­ date; in dry stigma such as Brassica, they are present in the pollen coat. Lack of suitable lipids on the surface of the stigma or pollen coat may act as a barrier for pollen tube entry. The stage at which the barriers operate seems to be related to the evolutionary diver­ gence of the parental species. When the stigma of Gladiolus (Iridaceae) was pollinated with the pollen of Crocosmia, belonging to the same fam­ ily, pollen hydration and germination were nor­ mal but pollen tubes failed to enter the stigma (Knox et al. 1976). However, when the pollen of Gloriosa, belonging to a different family (Liliaceae) was used for pollination, pollen hy­ dration itself was inhibited. Failure of pollen tubes to reach the ovary (af­ ter entering the stigma) is another most frequent prefertilization barrier. This is due to cessation of pollen tubes in the stigma, at various levels in the style or slow growth of pollen tubes, with the

174 Pollen Biology and Biotechnology

result that the pistil commences senescence before the pollen tubes reach the ovary. Such arrested tubes show many abnormalities, e.g. thicker walls, excessive deposition of callose, swollen tips and branching of the tube (de Nettancourt 1977, Dumas and Knox 1983, Fritz and Hannaman Jr 1989, Barone et al. 1992). It is now well established that pollen tubes grow­ ing in the pistil utilize stylar nutrients (see Chap­ ter 8). Inability of pollen tubes to utilize stylar nutrients may be a cause for pollen tube arrest. This may be due to the absence of suitable nu­ trients in the pistil or of suitable enzyme in the pollen to utilize the available stylar nutrients. Recent studies have shown that extracellular matrix of the transmitting tract plays a crucial role in pollen-pistil interaction (see Chapter 8). Considerable data indicate that some of the pro­ teins synthesized exclusively in the transmitting tissue (TTS proteins) participate in pollen tube growth, nutrition and their guidance. Absence of specific TTS proteins required to facilitate pollen tube growth may act as an important bar­ rier. Many crosses involve parents that show sig­ nificant difference in length of style; in some, such as species of Nicotiana, Datura and Lilium, this difference may be up to several centimetres. Crosses in which a short-styled type was used as the female parent and long-styled as the male parent were often successful while reciprocal crosses were not (Avery et al. 1959, Maheshwari and Rangaswamy 1965, Gopinathan et al. 1986, Potts et al. 1987). In maize x sorghum cross, pollen tubes do not reach the ovary because of the unusual length of maize silk. However, when pollen grains were deposited on the basal parts of the silk, pollen tubes readily reached the ovary (Heslop-Harrison et al. 1985a, b). In such crosses failure of the pollen tube to complete growth through the style appears to be due to the intrinsic inability of pollen tubes of the male parent to grow beyond the length of its own pis­ til, rather than to active inhibition. Examples of interspecific crosses showing inhibition of pollen tubes in the ovary are lim­

ited. In crosses involving many species of Poaceae, pollen tubes grow normally and reach the ovary. In maize x sorghum cross, no barri­ ers were observed until pollen tubes entered the ovary; however, the pollen tube failed to enter the ovule (Heslop-Harrison et al. 1984, 1985a, b). As entry of the pollen tube into the ovule is considered to be mediated through se­ cretion of a chemotropic substance in the micropyle, absence of such an attraction may act as a barrier. Recent studies have shown that the synergids provide guidance cues for pollen tubes to change the direction towards the ovule (see Chapter 8). Absence of such cues may form a barrier preventing entry of the pollen tube into the ovules. Even after entry of the pollen tubes into the embryo sac, many crosses show disturbances in double fertilization. In the wheat x maize cross, only 12% of the ovules showed development of both the embryo and endosperm, 80% of the ovules showed only embryo development but not the endosperm, and the remaining 8% showed only endosperm development but not the em­ bryo (Laurie and Bennett 1988). Obviously, this is due to some limitation(s) for effective syngamy and triple fusion; the controlling factors after dis­ charge of male gametes in the embryo sac are yet to be understood. In many crosses, prefertilization barriers are not restricted to a particular level or event, but are active over most of the post-pollination events. Thus the proportion of pollen grains that completes sequential post-pollination events is reduced at each level (Fritz and Hannaman 1989) with the result that very few or no pollen tubes reach the embryo sac. In the cross Vigna unguiculata x V. vexillata (Barone et al. 1992), 6.1% of the pistils showed inhibition of pollen germination, 5.3% showed pollen tube inhibition on the stigma surface, 32.6% showed inhibition inside the stigma and only in 8 .2% of the pistils did the pollen tubes reached the base of the style. As against 76.6% of the ovules showing fertilization following intraspecific compatible pollinations, only 18.3% ovules showed

Interspecific Incompatibility 175

fertilization in this cross. Similarly, in the cross Brassica fruticulosa x B. campestris, only 30% of the ovules showed pollen tube entry in con­ trast to 80% of the ovules in compatible pollina­ tions (Nanda Kumar et al. 1988). In many species different accessions show variation in intensity of prefertilization barriers (Snape et al. 1979, Zeven and Keijzer 1980). In the cross Vigna vexillata x V. unguiculata, for example, acc. TVnu 72 as the male parent re­ sulted in fertilization of 18.6% of the ovules, while acc. TVnu 73 resulted in fertilization of 31.7% of the ovules (Barone et al. 1992). Of over a dozen accessions of Tripsacum dactyloides used in crosses with maize, only one proved effective in achieving fertilization (Harlan and deWet 1977).

POST-FERTILIZATION BARRIERS In crosses showing post-fertilization barriers, pollen germination, pollen tube entry into the stigma, pollen tube growth through the style and pollen tube entry into the ovule proceed normally (Fig. 10.2D-F). Post-fertilization barriers result in the abortion of fertilized ovules at different stages of development. This is due to the break­ down of the endosperm and/or embryo. Barri­ ers may operate even during germination of the hybrid seed and subsequent growth of the F, hybrid (Zhou et al. 1991, Marubashi et al. 1988). Post-fertilization barriers are more prevalent than prefertilization barriers. Most crosses that show

prefertilization barriers show post-fertilization barriers also. Details of post-fertilization barri­ ers are much less understood. On the basis of indirect evidence many causative factors have been suggested. Some of these are: presence of lethal genes, genic disharmony in the embryo, failure of the endosperm to develop or its early breakdown, and unfavourable interactions be­ tween the embryo sac and the surrounding ovu­ lar tissue (Khush and Brar 1992). In several crosses embryo abortion is associated with pro­ liferation of nucellar or integumentary cells (Cooperand Brink 1940). Chromosome elimination, in which chromo­ somes of one of the parents get eliminated dur­ ing embryo development, is another manifesta­ tion of post-fertilization incompatibility, particu­ larly in cereals (see Chapter 14). Chromosome elimination leads to the development of haploids; in most of the crosses embryo rescue is needed to recover haploids (see Palmer and Keller 1997). Many instances of haploid production in interspecific crosses, which are generally con­ sidered to be due to parthenogenesis, may be the result of chromosome elimination. Detailed cytological studies are needed to understand post-fertilization barriers in such crosses. De­ tailed discussions on post-fertilization barriers are presented by Stebbins (1958), Maheshwari and Rangaswamy (1965), de Nettancourt (1977, 2001 ), Hadley and Openshaw (1980), Raghavan (1986), and Khush and Brar (1992).


Pollen biotechnology is the management or manipulation of pollen grains for production and improvement of crops and other related eco­ nomic products. Pollen grains can be manipu­ lated during all phases of pollen biology—de­ velopment, free-dispersed phase, pollination, pollen-pistil interaction and fertilization. Impres­ sive progress made in recent years on pollen biology, covered in earlier chapters, has made the applications of pollen biotechnology more diverse, effective and viable. A steady and significant increment in crop productivity has become imperative to feed the growing millions. There are two possible strategies to enhance crop productivity: (i) realization of potential yield by providing optimal conditions for plant growth and (ii) enhancement of genetic potential of crop species through transfer of agronomic or other useful traits from other species/genera. Increase in productivity through both these strategies has to be achieved under many constraints, some of which are: • non-availability of additional land for culti­ vation, • steady decrease in area of agricultural land due to increasing salinity, alkalinity, water­ logging, soil erosion and industrial contami­ nation and • need to drastically reduce the application of environment-unfriendly agrochemicals such as herbicides and pesticides. Future plant breeders have therefore to de­ velop varieties that give not only higher yield with better quality, but are also able to grow on mar­ ginal lands with minimum use of environmentunfriendly chemicals. Realization of these objectives is very chal­ lenging and traditional methods of crop produc­ tion and improvement are inadequate for achiev­ ing them. Some of the major challenges that cannot be effectively tackled by traditional meth­ ods are:

• insufficient pollination, particularly in many cross-pollinated species, due to lack of ad­ equate number of pollinators; • inability to exploit hybrid vigour in a number of crops due to non-availability of effective pollination control systems, • narrow genetic base of crop species which has greatly increased their susceptibility to biotic and abiotic stresses such as diseases, pests, drought and salinity; and • need to extend breeding programmes be­ yond species limits since the required genes are no longer available in the cultivated spe­ cies. Constant infiltration of new adaptive genes into cultivars is necessary to sustain and fur­ ther improve the yield. Many wild and weedy relatives of crop species provide a vast source of such genes. However, strong crossability bar­ riers between wild and cultivated species make gene transfer almost impossible using traditional methods. Integration of pollen biotechnology into the traditional breeding programme offers effective approaches in a number of areas of production and improvement of crops and other products. Like any other technology, pollen biotechnology also decreases the time and cost of crop im­ provement and increases its efficiency. The fol­ lowing are the potential areas of pollen biotech­ nology presently available for effective applica­ tion: • overcoming pollination constraints, • developing effective pollination control sys­ tems to exploit hybrid vigour, • overcoming crossability barriers and many other constraints to transfer genes across species barriers and • developing other economic products related to pollen. The chapters on pollen biotechnology give a concise and integrated account on the progress made in these areas.

11 Optimization of Crop Yield Fruits and seeds are the economic products of most crop plants. Effective pollination is a pre­ requisite for fruit- and seed-set. Therefore, suc­ cessful pollination is of vital importance to real­ ize optimal yield. In self-incompatible species, yield is dependent on adequate cross-pollina­ tion. Even in self-compatible species, pollination is largely dependent on pollinating agents as automatic selfing seldom occurs or is insufficient in most of the self-compatible crop species. The majority of our crop plants, except cereals (which are anemophilous), are pollinated by insects particularly bees. Adequate pollination is often a major constraint in many crop species due to one or more of the following reasons: 1. Drastic reduction in native pollinator popu­ lations because of the steady disappear­ ance of natural habitats of insects, marked increase in level of pollutants and extensive use of environment-unfriendly agrochemi­ cals, in particular pesticides and herbicides. 2. Lack of sufficient number of native pollina­ tors due to enormous increase in area cov­ ered by the same crop species (often ex­ tending to hundreds of hectares) in the present-day monoculture cropping system. 3. Absence of natural pollinators for crops in­ troduced from other regions. Over 60 years ago it became apparent in the USA that production in many of the fruit, seed and nut crops could be increased substantially

by careful management of pollination (Roubik 1995). This led to the initiation of extensive stud­ ies to increase pollination efficiency in crop spe­ cies. Increased pollination efficiency can lead to increase in crop value by increasing crop yield, uniformity, quality and decreasing the time of crop maturity. Several effective approaches are available to overcome pollination constraints (Jay 1986, Torchio 1990, Free 1993, Currie 1997). However, progress in pollination management is largely confined to developed countries. In de­ veloping countries, although the yield in a num­ ber of crop species has been identified to be pollination-limited, pollination management has not yet been integrated into regular crop pro­ duction practices (Savoor 1998), largely be­ cause of lack of data on pollination biology, pol­ lination efficiency and details of pollinators of various crops (Roubik 1995).

ENHANCING POPULATIONS OF NATIVE POLLINATORS THROUGH HABITAT MANAGEMENT Increasing local populations of native pollinator species through habitat management is one approach to removing pollination constraint. This is particularly useful when availability of nest site is a limiting factor. Habitat management can be achieved by maintaining uncultivated strips along field margins and providing permanent

182 Pollen Biology and Biotechnology

nest boxes (Pomeroy 1981, Osborne et al. 1991). Such strategies should also ensure availability of forage sources when the target crop is not in bloom.This approach also requires management of cropping practices in such a way that the flow­ ering period of the target crop coincides with the peak populations of the pollinator (Free 1993). However, management of habitat is more expen­ sive, particularly in areas of intensive agronomic practices (Torchio 1990).

USE OF COMMERCIALLY MANAGED POLLINATORS So far the most economically viable and effec­ tive approach to overcome pollination constraints has been use of commercially managed polli­ nators, in particular honey-bees (Apis spp.), for pollination services. Honey-bees are the most effective pollinators in a range of crop species (legumes, fruits, vegetables and nuts) (Robinson 1979, Schmidt 1982). Management of honey­ bees is convenient because of their large forag­ ing populations, year-round availability and easy transportation. Methods for their management and control of their diseases, predators and parasites are well established (van Heemert et al. 1990, Free 1993). Increasing pollination effi­ ciency through management of pollinators is warranted only when crops are pollen-limited and the cost involved is lower than the value realized through increase in crop production (Currie 1997). The acreage of bee-pollinated crops has in­ creased enormously in recent years. It is esti­ mated that the value of bee-pollinated crops in the USA alone is ca US $9.3 billion.The demand for pollinators is steadily increasing (Torchio 1990). In India also many studies have indicated significant increase in yield of several oil-seed crops, especially sunflower, as a result of intro­ duction of honey-bee colonies (Anonymous 1996, Savoor 1998). Introduction of beehives in crop fields to en­ hance pollination efficiency and thus crop yield was known as early as 1892 (Savoor 1998).This

technology was introduced on a commercial scale in the USA during the 1940s. Use of honey­ bees for pollination services has grown steadily over the years and it is presently estimated that in the USA alone over one million honey-bee colonies are rented every year for pollination services (Torchio 1990). These colonies are moved from one crop to another coinciding with the flowering season. Hives rentals in the last 2-3 years in the USA is reported to be ca $45 per colony/crop season (Gadagkar 1998). In recent years, maintenance of honey-bee colo­ nies for pollination services has become more profitable and honey and wax have become by­ products. Until recently, management of honey-bees for pollination services was confined largely to fieldgrown crops. Lately honey-bees have been in­ creasingly used to achieve pollination in glass­ house crops, particularly tomatoes. Apart from bees, many other pollinators have shown po­ tential for use in pollination services (Torchio 1987). For some crops, bumble-bees (Bombus spp.) are better pollinators (Holm 1966) than honey-bees because of their larger size, long tongue, and ability to vibrate flowers and fly at relatively low temperatures (Currie 1997). How­ ever, the high cost associated with bumble-bee management has so far restricted their use largely to red clover (Macfarlane et al. 1983) and to high-value glasshouse crops (Banda and Paxton 1991, Straver and Plowright 1991, Kearns and Inouye 1993). A commercial com­ pany in Europe is reported to rear >1000 bumble bee colonies per year for pollination services of glasshouse-grown tomato (Kearns and Inouye 1993). Pollen collectors are generally more effec­ tive pollinators than honey collectors. Because pollen is primarily required for providing food for larvae and young bees, the proportion of pollen collecting bees is correlated with the level of egglaying by the queen and the amount of brood present in the colony. Supply of extra frames of brood to colonies generally increases pollen collection.

Optimization of Crop Yield 183

Pollen transfer can be made more efficient by using ‘pollen dispensers’ mounted in the hive. The foragers are forced to pass through these dispensers which contain pollen of a compat­ ible cultivar and thus pollen grains get depos­ ited on the body parts of the foragers (Legge 1976, Ferrari 1990). Some of this pollen is even­ tually transferred to the stigmas of the crop plants during forager visits (Griggs and Iwakiri 1960, Williams et al. 1979). Use of pollen dispensers may prove particularly valuable in the produc­ tion of hybrid seeds since pollination of male sterile plants can be achieved even in the ab­ sence of male fertile plants (Farkas and Frank 1982). This technology is likewise useful for self­ incompatible species, in particular orchard spe­ cies for which compatible trees are absent or insufficient. Many private companies sell pollen of different crop species for use in pollination services (Kearns and Inouye 1993, Newman 1984). The pollen in dispensers has to be con­ tinually replaced. The required number of honey-bee colonies is maintained close to the target crop. Timing of pollinator introduction should synchronize with flowering of the target crop; introduction of bees too early or too late would reduce overall polli­ nation efficiency. During inclement weather, more colonies are required to achieve pollination. In­ formation is available for a number of crops de­ tailing the requirement for populations of honey­ bees/bumble-bees to pollinate a given area and their management In the field (DeGrandiHoffman 1987, Free 1993). Use of honey-bee colonies for pollination services poses some management problems; honey-bees forage on a wide range of host plants and thus maintaining them on the target crop often becomes difficult. Thorough knowl­ edge is required about the behaviour of bees, pollination biology of crop species and compet­ ing crops, and the ecology of plant-pollinator relationship. Effective management strategies have to be followed to retain pollinators on the target crop. Honey-bees locate flowers by sight and odour

and their fidelity to the crop is determined by the quantity and quality of the reward the crop offers. In the absence of sufficient rewards, bees desert commercial target crops for other attrac­ tive pollen- and nectar-yielding species (com­ peting species) (Jay 1986). When the target crop is sprayed with an insecticide, bee repellents are sprayed on the target crop to temporarily pre­ vent bee visits to the crop species, thereby re­ ducing bee mortality. Bees generally expand their foraging range gradually after being released from colonies at a new sight. When their foraging extends beyond the target crop, the colonies need to be rotated or replaced with fresh ones. This management strategy is particularly effective when the flow­ ers of the target crop are not very attractive to the bees. Overlapping of the flowering period of the target crop with a competing crop can be avoided by changing the planting period of the target crop. When weeds divert the pollinators from the target crop, competition from weeds can be re­ duced by mowing them or treating them with herbicides. Spraying competing plants with bee repellents such as carbolic acid, acetic acid, benzaldehyde or calcium chloride is another approach to reducing bee visits to competing non-target crops (Jay 1986). One of the most effective long-term strate­ gies for improving pollination efficiency is ratio­ nal breeding of crops as well as honey-bees. Breeders should be able to breed crop varieties to make them more attractive and to provide better rewards in the form of more nectar pro­ duction (see Currie 1997). Similarly, breeding bees that have a limited or reduced flight range and show preferential attraction to specific crops would be a great advantage in retaining bees on target crops (see Jay 1986).

Spraying Pollinator Attractants on Target Crop A number of investigations have shown the po­ tential of sprays with various substances on the target crop to attract pollinators (see Jay 1986,

184 Pollen Biology and Biotechnology

Currie 1997). Although spray with sugar syrup often increases the abundance of foragers on the target crop, it may not increase crop yield, as the foragers tend to collect the syrup rather than the pollen or nectar. Sprays with dilute solutions of pheromones (chemicals used for communication between members of the same species) have shown con­ siderable potential. Secretions from the Nasonov gland located on the dorsal side of the abdo­ men of worker bees consist of seven terpenoids. Of these, geraniol and citral have been shown to increase honey-bee foraging activity. Similarly, the queen honey-bee secretes a five-component pheromone from its mandibular glands. Sprays of synthetic mandibular pheromones have been shown to increase honey-bee foraging activity and crop yield under a wide range of conditions (Currie et al. 1992a,b; Winston and Slessor 1993). Application of pheromone seems to be particularly effective on crops with flowers rela­ tively unattractive to bees or during inclement weather conditions. Apart from pheromones, sprays of synthetic plant volatiles isolated from nectar or pollen are also effective in attracting honey-bees (see Dob­ son 1994). Sprays containing food supplement such as Beeline(R) have also been reported to act as bee attractant in some crop species (Margalith etal. 1984).

INTRODUCTION OF POLLINATORS When crops are grown in areas where natural pollinators are absent, as often happens when a crop is introduced from one country to another, introduction of pollinators is one of effective ap­ proaches. This approach involves detailed stud­ ies on the biology of the pollinator and monitor­ ing its establishment in the new area. Oil palm (Elaeis guineensis) is a native of Africa and Central South America. It was introduced to Malaysia and Indonesia at the beginning of this century, where it is grown extensively at present. In its native habitat, oil palm is pollinated by wind as well as many insects in particular weevils. In

many parts of Malaysia, where pollinating in­ sects are absent, natural pollination was inad­ equate. Introduction of the weevil Elaeidobius kamerunicus, an important pollinator of oil palm from Cameroon, to Malaysia has been a suc­ cessful example of such an approach. Introduc­ tion of weevil during the 1980s has markedly increased the yield in these areas (Syed 1979). Weevil could establish well in oil-palm estates in most Malaysian plantations. Over the first seven months after weevil introduction, oil yield increased 20-53% (Syed 1979, Syed et al. 1982).

SUPPLEMENTARY POLLINATION/ ASSISTED POLLINATION During prolonged inclement weather conditions insufficient pollination is common in many crops in spite of availability of a sufficient number of pollinators. Many orchard species, such as apple, pear and cherry, are self-incompatible and are clonally propagated. Fruit-set depends on availability of two or more intercompatible vari­ eties. Even when such lines are available, the activity of bees is greatly reduced in cold and wet weather, resulting in insufficient pollination. Raising of clonally propagated orchards using self-compatible clones would not only overcome pollination constraints, but also permit raising orchards with a single elite variety (see Reimann-Phillipp 1965). It is possible to induce self-compatibility through induction of mutations in SI alleles by irradiation of flower buds (Lewis 1954). However, this approach cannot be used for the existing orchards. Assisted/supplemen­ tary pollination through pollen sprays (Brown and Perkins 1969, Williams and Legge 1979, Hop­ ping and Jerram 1980a, b) or any other method is the most effective technique for sustaining crop yield.This is routinely carried out for small-scale production of a few crops such as passion-fruit (Roubik 1995). In high-value plantation crops such as oilpalm, pollination is a major constraint even in the presence of weevils, especially in younger

Optimization of Crop Yield 185

plantations. Although oil palm is monoecious, the male and female phases alternate, each extend­ ing for many months; thus at any given time the plant will be in either the male or female phase. Insufficient number of plants in the male phase in the plantation and unfavourable weather con­ ditions (Hardon and Turner 1967, Veldhuis 1968) reduce pollination efficiency. Assisted pollination is a common practice in oil-palm plantations in Malaysia and Indonesia, particularly in younger plantations (Hartley 1988). Assisted pollination has been reported to increase yield 20-150%, depending on age of the plants and weather conditions (Hardon 1973). Assisted pollination requires standardization of protocols for pollen collection, pollen storage and pollination. Different methods have been essayed for assisting pollination in oil-palm (Wong and Hardon 1971). For pollinating young

palms a ‘hand puffer’ (a rubber bulb/plastic bottle with an extension tube) has been found conve­ nient; pollen is applied directly on anthesizing female inflorescences. For taller palms, blanket dusting of crowns is practised using a portable mechanical blower or duster (Wong and Hardon 1971). Pollination is normally repeated at 3-day intervals. Mechanized pollination of fruit trees such as apple and peach by using sprays has been tried, but this approach has not yet reached a commercial scale. Hand pollination is regularly practised for Vanilla orchid. This orchid is a native of south­ ern Mexico and Central America where it is pol­ linated by the euglossine bee Eulaema (Roubik 1995). Vanilla is grown extensively in many parts of tropical Asia where the pollinator is absent; hand pollination is routinely carried out to induce fruit-set.

12 Commercial Production of Hybrid Seeds Development of hybrid seed technology for im­ provement of crop productivity is one of the most important advances in agriculture. The presence of sufficient heterosis/hybrid vigour is the primary requirement for application of hybrid seed tech­ nology. Hybrid vigour refers to an increase in vigour and productivity of the hybrid compared to its parents. Hybrids also show greater unifor­ mity. For commercial purposes the extent of hy­ brid vigour is assessed on the basis of high par­ ent heterosis (yield of F, hybrid exceeding the better parent used in the cross) as well as stan­ dard heterosis (yield of F, hybrid exceeding the best non-hybrid cultivar). The basis of the hybrid seed industry was established during the second half of the 19th century when several geneticists showed that varietal hybrids of corn were more vigorous than inbred varieties (see Allard 1960). It was Shull who suggested at the beginning of the 20th century, use of inbred lines obtained by continu­ ous self-pollination for commercial production of hybrid seeds (single-cross hybrids). However, this method did not lead to commercial use of hybrid varieties due to lack of suitable inbred lines and many other limitations (see Allard 1960). The double-cross method, suggested by Jones in 1918, made commercial production of hybrid maize economically feasible. Although the first commercial production of hybrid maize was

initiated in 1921, its impact was felt in agricul­ ture only from the 1930s onwards. In 1933 hy­ brid maize covered • XA

Production of stock seeds by growing line XA and XB / YA and YB in alternate rows




S,S 2°2

S 3S 3


s 4s 4





S iS 2

S 3S4

Production of hybrid seeds by growing XAB and YAB lines in alternate rows Hybrid seeds Fig. 12.4 Major steps involved in hybrid seed production using self-incompatibility. S,, S2, S3 and S4 alleles indicate SI geno­ types.

particular brassicas—cabbage, cauliflower, kales and Brussels sprouts (Odland and Noll 1950, Thompson 1978, Gowers 1980, Pearson 1983). The following are some of the limitations for use of SI lines in hybrid seed production. 1. SI poses special problems in developing and maintaining pure lines (Fig. 12.4). Development of pure lines and their maintenance requires repeated self-pollina­ tion. As SI does not permit seed-set in self­ pollinated flowers, some effective technique has to be used to overcome SI. Standard­ ization of an effective and simple protocol to overcome SI is an essential requirement for developing and maintaining pure lines. This is cumbersome and expensive (see below), and often pure lines show inbreed­ ing depression.

2. Susceptibility of SI to environmental varia­ tions, in particular temperature and humid­ ity. This results in contamination of hybrid seeds with selfed seeds to different degrees depending on the prevailing environmental conditions. 3. SI is not present in many crop species. Many approaches are in use or have the poten­ tial to circumvent these limitations.

Methods to Overcome SI A range of methods is available to overcome SI (Table 12.1). However, most are cumbersome and can only be applied on a limited scale. This greatly increases the cost of hybrid seed pro­ duction. Bud pollination has been the most com­ mon method used so far to overcome SI for de­ velopment and maintenance of pure lines


Pollen Biology and Biotechnology

Table 12.1 Some of the effective techniques to overcome self-incompatibility Technique



Induction of mutations

Oenothera Prunus Pétunia Trifolium Nicotiana

Lewis 1949 Lewis and Crowe 1953 Brewbaker and Natarajan 1960 Denward 1963 Pandey 1965, van Gastel and de Nettancourt 1975

Induction of autotetraploidy

Pyrus Pétunia Lolium

Crane and Lewis 1942 Stout and Chandler 1941 Ahloowalia 1973

Bud pollination

Brassica Raphanus Pétunia

Attia 1950, Odland and Noll 1950 Kakizaki 1930 Eyster 1941, Shivanna and Rangaswamy 1969

Delayed pollination

Brassica Lilium

Kakizaki 1930 Ascherand Peloquin 1966a

Hot water/high temperature treatment

Malus, Pyrus, Prunus Oenothera Trifolium Brassica Lycopersicon Lilium Raphanus

Lewis 1942 Hecht1961 Kendall and Taylor 1969 El Muraba 1957 Hogenboom1972 Ascher and Peloquin 1966b Matsubara 1980

Application of growth substances


Emsweller and Staurt 1948 Emsweller et al. 1960 Matsubara 1973 Eyster 1941

Pétunia, Trifolium, Tagetes, Brassica Use of mentor pollen

Theobroma Cosmos Malus Pétunia Nicotiana

Opeke and Jacob 1969 Howlettet al. 1975 Dayton 1974 Sastri and Shivanna 1976a Pandey 1979

Placental pollination


Rangaswamy and Shivanna 1967 Niimi 1970, Wagner and Hess 1973

High carbon dioxide concentration (3-6%)


Nakanishi and Hinata 1975, Dhaliwal et al. 1981, Taylor 1982, O’Neill et al. 1984, Vitanova 1984, Tao and Yang 1986

Treatment of flowers with high C 02 and humidity


Gowers 1980

Brassica Treatment of stigma with NaCI (1.5-3%) before or after pollination Treatment of stigma with lectins or pollen with sugars before pollination

Petunia, Eruca

(Pearson 1983). It Is effective in a range of spe­ cies. However, it is labour intensive since it re­ quires handling individual flowers. There is a need for development of simpler and more ef­ fective technique(s) to overcome SI.

Tao and Yang 1986, Monteiro et al. 1988, Fu et al. 1992 Sharma et al. 1985

Treatment with high C 0 2 concentration or sodium chloride is very promising for overcom­ ing SI in species of Brassica. C 0 2treatment can be given only in closed glasshouses and thus only a limited amount of seeds can be harvested.

Commercial Production of Hybrid Seeds 193

Treatment with sodium chloride is more conve­ nient and much less expensive. Since it can be applied under field conditions, a large amount of seeds can be harvested. However, the effi­ cacy of both these treatments is variable be­ tween species and genotypes, and is affected by prevailing environmental conditions. There­ fore, detailed studies are needed to standard­ ize the protocol for different genotypes. Apart from its use in the production of hybrid seeds, effective methods to overcome SI are necessary for other practical purposes. For ex­ ample, selfing is the only alternative method to maintain hybrid characters in ornamental spe­ cies which are difficult to propagate vegetatively or prone to viral infection. For genetic and breed­ ing studies also it is desirable to obtain homozy­ gous progeny. SI is a hindrance for such studies. Many orchard species, e.g. apple, pear and cherry, are self-incompatible and must be clonally propagated. Fruit-set depends on the availability of two or more intercompatible lines. Even when such lines are available, bee activity is greatly reduced in cold and wet weather con­ ditions, resulting in insufficient pollination. Rais­ ing clonally propagated orchards using self-com­ patible clones would not only overcome pollina­ tion constraints, but also permit raising orchards with a single elite variety (see Reimann-Phillipp 1965). It is possible to induce self-compatibility through induction of mutations in SI alleles through irradiation of flower buds (Lewis 1954). However, this approach cannot be used in the existing orchards.

Selection of Lines with Strong SI Alleles To circumvent the problem of breakdown of SI under varied environmental conditions, it is nec­ essary to select genotypes with strong SI alle­ les. This can be done either on the basis of seedset in self-pollinated flowers or by scoring pol­ len tube penetration in selfed pistils through aniline blue fluorescence technique. A method which combines testing for the level of SI with production of selfed seeds through bud pollina­

tion, has been suggested for Brassica (Pearson 1983). Both open flowers and buds (3-4 days before anthesis) are self-pollinated in each in­ florescence; open flowers are separated from pollinated buds by a tag tied to the inflorescence axis. After fruit maturity, the number of seeds from each part is determined. An effective SI allele is indicated by 50% of the ovules into seeds in self-pollinated buds (ca 10-12 seeds/ pod) indicates the suitability of a line for pure line production through bud pollination.

Transfer of SI Alleles to Self-compatible Species Most of the crop species are self-compatible (SC) although many of their wild relatives are SI. Transfer of SI (from SI species) to SC culti­ va rs/species has not been feasible through con­ ventional breeding programmes because of many practical difficulties (de Nettancourt 2001 ). Recent advances in cloning and characteriza­ tion of S-genes (see Chapter 9) have opened up the possibility of transferring the S-gene through recombinant DNA technology.

HYBRID SEED PRODUCTION IN MONOECIOUS SPECIES In many members of Cucurbitaceae and in spin­ ach (Spinacea oleracea), which are mono­ ecious, hybrid seeds are produced through ma­ nipulation of the development of male and fe­ male flowers (Pearson 1983).The proportion of male and female flowers on each plant is highly variable depending on the genotype of the plant and on prevailing environmental conditions; this proportion can also be altered by application of growth substances. In Cucurbita pepo, male flowers which de­ velop early in the flowering period, are manu­ ally removed from the seed parent (in seed pro­ duction plots) at 2- or 3 -day intervals to prevent self-pollination.This procedure though effective, increases the cost of hybrid seeds; it also often results in contamination of seeds since rogueing

194 Pollen Biology and Biotechnology

of all male flowers is difficult to achieve. In re­ cent years a more convenient method based on application of plant growth regulator to induce development of only male or female flowers has been employed. Production of male flowers in the seed parent is suppressed for 2-3 weeks by spraying seedlings with ethylene. Female and male parents are usually planted in a 2:1 ratio in seed production plots, and the plants of the pollen parent are destroyed after the crop has set. This technology has markedly reduced the cost of hybrid seed production. In many other cucurbits (e.g. C. sativus and C. melo) also, auxins and ethylene induce fe­ male flowers while gibberellic acid and antiethylene substances such as silver nitrate induce male flowers. In these crops, highly gynoecious lines have been isolated for use in hybrid seed production. Homozygous gynoecious lines are maintained by using gib­ berellic acid or silver nitrate to induce male flow­ ers; self- and sib-pollinations in gynoecious plants are achieved through insects. Gynoecious and normal monoecious lines are raised in seed production plots and cross-pollination is achieved through insect activity. Spinach (Spinacea oleracea) is monoecious and wind pollinated. The extent of female flow­ ers on each plant ranges from 100% to 0%. Highly gynoecious female lines have been used for hybrid seed production. Unlike cucurbits, in which early-formed flowers are males, in spin­ ach early-formed flowers are females. In highly gynoecious female lines, male flowers are formed only when female flowers do not set seeds and thus fail to act as metabolic sinks. If sufficient pollen is available in the field, as is the situation in seed production plots, early-formed female flowers set seeds which act as sinks; the plant matures with its seed crop and senesces without producing male flowers.

Manipulation of Sex-expression for Yield Improvement In many dioecious (e.g. papaya) and monoe­ cious (many cucurbits) crop species, productiv­

ity can be increased by increasing the number of female plants (in dioecious species) or female flowers (in monoecious species). Following the demonstration that many growth substances shift the sex-expression in favour of male flow­ ers or female flowers such a technology has become feasible. Treatment of male plants with auxins or ethylene induces development of fe­ male flowers. Similarly treatment of monoecious plants with auxins or ethylene increases the number of female flowers. Such an increase in female flowers results in increased fruit- and seed-set.

USE OF POLLEN STERILITY INDUCED THROUGH r-DNA TECHNOLOGY Details of development of pollen sterile and re­ storer systems using r-DNA technology have been described in Chapter 5. This pollination control system is being perfected for commer­ cial production of hybrid seeds in Brassica napus. To overcome the problem of rogueing of male fertile plants in seed production plots, a herbicide resistant gene, 35S bar, has been in­ troduced into the female line along with TA-29 barnase (Fig. 12.5). Treatment of plants of the female parent with herbicide at the seedling stage eliminates male fertile plants, as they are susceptible to the herbicide. The transgenicbased pollination control system is undergoing field trials in Brassica napus and is likely to be released soon for commercial production of hy­ brid seeds.

USE OF POLLEN STERILITY INDUCED THROUGH CHEMICAL HYBRIDIZING AGENTS Considerable progress has been made in recent years in development of effective chemical hybridizing agents (CHAs), particularly for wheat (see Chapter 5). A few have shown potential for commercial application (Cross and Schulz 1997). Limited studies have also been carried

Commercial Production of Hybrid Seeds 195



male sterile + herbicide resistant

male fertile herbicide susceptible


+ 35S


male fertile + herbicide susceptible

male sterile + herbicide resistant



herbicide treatment at seedling stage (eliminates male fertile plants)

male fertile with restorer gene (TA29-barstar)

male sterile + herbicide resistant

male fertile hybrid Fig. 12.5 Diagram of hybrid seed production using barnase-barstar approach by linking TA-29 barnase with selectable marker gene 35S bar which confers tolerance to the herbicide glufosinate ammonium.

out to understand their mode of action. The following are some of the CHAs that have shown potential for application in hybrid seed production.

Proiine Analogues Proline is the most abundant free amino acid in pollen grains. Accumulation of high concentra­ tion of proline seems to be necessary for the production of fertile pollen. A few proline ana­ logues such as methanoproline (c/s-3,4-methylene-S-proline), and azetidine-3-carboxylic acid (A3C) have been reported to be effective in in­ ducing pollen sterility in wheat and barley (Devlin and Kerr 1979, Searle and Day 1980, Porter et

al. 1985, Johnson and Lucken 1986). Although pollen grains in A3C-treated plants appeared normal in morphology, they produced only small pollen tubes on the stigma, which eventually burst or failed to enter the stigma (Ladyman and Mogensen 1987, Mogensen and Ladyman 1989). However, these pollen grains surprisingly showed higher levels of proline accumulation indicating that A3C may not act as a proline ana­ logue, but affect pollen development in some other way (Mogenson and Ladyman 1989).

Phenyl Pyridazones Fenridazon (RH-0007), a phenyl-substituted pyridazone carboxylate developed by Rohm and

196 Pollen Biology and Biotechnology

Hass, has been reported as an effective CHA, allowing commercially acceptable level of seed yield (Bucholtz 1988). This chemical did not af­ fect meiosis in sporogenous cells, but affected development of microspores (Mizelle et al. 1989). The microspores showed irregular or reduced exine and eventually aborted. The tapetal cells enlarged and protruded into the locule; the orbicules deposited on the inner tapetal wall were greatly reduced. The pollen wall was 80% thinner in treated plants (El-Ghazaly and Jensen 1986). It was suggested that RH 0007 inhibited polymerization of sporopollenin precursors into exine wall and orbicules (Mizelle et al. 1989).

Phenylcinnoline Carboxylates These CHAs have been developed by Sogetal Inc. (France) (Guilford et al. 1992). SC 1058 [ 1(4 trifluoromethyl phenyl)-4-oxo-5-fluorocinnoline-3-carboxylic acid] and SC 1271 [1-(4‘chlorophenyl)-4-oxo-5-propoxy-cinnoline-3-carboxylic acid] induce complete pollen sterility in wheat when applied at the premeiotic stage of anther development (Cross et al., 1989,1992). The effects of these CHAs are very similar to fenridazon (RH0007) (Schulz et al. 1993), indi­ cating that SC 1058 and SC 1271 also mediate pollen sterility by interfering with the supply of sporopollenin wall materials (Schulz et al. 1993). Subsequently, another compound, SC 2053 [1-(4'-chlorophenyl)-4-oxo-5-methoxyethoxycinnoline-3 -carboxylic acid] was tested for CHA activity in wheat (Batreau et al. 1991).Treatment of plants with SC 2053 (700 and 1000 g/ha), when the spikes were 11-20 mm long, resulted in a high degree of pollen sterility (Batreau et al. 1991, Patterson et al. 1991, Cross and Schulz 1997). These CHAs have been reported to be effective in producing a usable level of male ste­ rility with minimum phytotoxicity and loss of seed yield, when applied at the right stage and dos­ age.

LY 195259 LY 195259 [5-(aminocarbonyl)-1-(3-methylphenyl)-1H-pyrazole-4-carboxylic acid], a syn­

thetic growth regulator, has been reported to be a commercially promising CHA for wheat (Tschabold et al. 1988). Even low dosage (1.12 kg/ha) produced >95% pollen sterility with a high degree of female fertility (seed-set as high as 80%). No other major cytotoxic effects have been reported with this CHA. Administration of the chemical to the soil was also effective in induc­ ing pollen sterility (see Cross and Schulz 1997).

MON 21200 Monsanto reported the effects of a series of py­ ridine monocarboxylic and benzoic acid ana­ logues on CHA activity in wheat (Ciha and Ruminski 1991). Currently a Monsanto-affiliated company (Hybri Tech) is developing a CHA com­ pound MON 21200 for wheat (see Cross and Schulz 1997). MON 21200 is reported to pro­ vide good CHA activity over a wide range of genotypes, geographic regions and growth con­ ditions. Outcrossing rates of around 80% have been reported for different wheat varieties over 6/7-year field trials.

POTENTIAL OF APOMIXIS IN HYBRID SEED TECHNOLOGY One of the main constraints in hybrid seed tech­ nology at present is the need to produce hybrid seeds every year. Apomixis, an asexual repro­ ductive trait, has the potential to overcome this major limitation (Khush 1994, Hanna and Bashaw 1987, Hanna 1991, Spillane et al. 2001). Broadly, apomixis includes all forms of asexual reproduction (Nogler 1984, Koltunow 1993).The type of apomixis relevant to hybrid seed tech­ nology is the one in which seeds develop with­ out fertilization. Apomixis has been known for many decades but largely remained a trait of academic interest (Maheshwari 1950, Johri 1984). Apomixis has been reported in over 300 species belonging to about 40 families. It is more common in polyploid species, presumably as an adaptation to overcome meiotic abnormalities in seed development. Apomixis is prevalent in members of Asteraceae (Hieracium), Poaceae

Commercial Production of Hybrid Seeds 197

(Panicum, Pennisetum) and a few other fami­ lies. Several species are obligate apomicts (e.g. Commiphora wightii, Gupta et al. 1996) while many show facultative apomixis, producing seeds both apomictically and sexually. In most plants apomictic embryos originate either from an unreduced egg or from the sporophytic cells surrounding the embryo sac. An unreduced egg develops as a result of the fail­ ure of female meiosis (Antennaria) or formation of restitution nucleus after the megaspore mother cell enters meiosis (Taraxacum); both events result in formation of a 2 n embryo sac containing a 2n egg. In the other pathway the haploid egg of the reduced embryo sac degen­ erates but a sporophytic cell either from the nucellus or the endothelium (innermost layer of the integument) develops into the embryo. The former pathway is more prevalent. In many spe­ cies, such as Citrus, both sexual and apomictic embryos develop in the seeds; the reduced egg gets fertilized and develops into the sexual em­ bryo at the micropylar part of the embryo sac; apomictic embryos arise from the nucellar cells, protrude into the embryo sac and continue de­ velopment. This results in polyembryony, one of the embryos being sexual and the remainder apomictic. For full development of apomictic embryos, development of endosperm is necessary for nourishing the embryo. Endosperm development in many apomictic species is spontaneous (from fused polar nuclei) or pseudogamous (one of the sperms fuses with the secondary nucleus but the other sperm does not fuse with the egg). For the latter, pollination and pollen-pistii inter­ action have to be completed. Endosperm devel­ opment is necessary not only for embryo nour­ ishment, but also for human consumption in spe­ cies with endospermous seeds. In many crop species, especially in all cereals, the economi­ cally important part of the seed is the endosperm as it stores most of the nutrients. In the absence of endosperm development, apomixis is of no commercial use in such species.

Since meiosis and fertilization are not in­ volved in the formation of apomictic seeds, the progeny raised from such seeds will be identi­ cal to the parent. Transfer of the apomictic trait into the hybrid plants would result in fixation of hybridity; since the progeny do not show segre­ gation for hybrid vigour, the farmer can use the seeds collected from hybrids generation after generation. This would enormously simplify hy­ brid seed technology and increase profitability to the farmer. Parthenogenesis (development of embryo from reduced egg without fertilization), another form of apomixis, is not useful for hy­ brid seed technology as it involves meiotic seg­ regation and also results in sterile progeny. Realization of the potential of apomixis in hybrid seed technology has stimulated great in­ terest. Extensive studies are underway world­ wide for screening a large germplasm for apo­ mixis, understanding its genetics, and develop­ ing effective methods for its transfer to sexual crop species. Apomixis involves a series of events: failure of female meiosis, organization of the embryo sac, stimulation of the egg to de­ velop into the embryo without fertilization and development of endosperm. In instances of ad­ ventive embryony, it also involves stimulation of the nucellar or integumentary cells to develop into embryos. Apomixis therefore seems to in­ volve a number of genes controlling different aspects of meiosis and seed development; this makes transfer of the apomictic trait from one species to another more difficult. However, many studies have indicated that in several species, apomixis is controlled by a single gene or a few genes (Nogler 1984). The single loeus is likely to be comprised of many tightly linked genes. Genes controlling apomixis may be dominant or recessive. Although most of the crop species are sexual, many of their wild relatives show apomixis. Several attempts have been made to transfer apomixis from these sources to sexual species (Khush 1994) using traditional breeding approaches; for example, from Tripsacum

198 Pollen Biology and Biotechnology

dactyloides to maize, from Pennisetum orientale to pearl millet, from Agropyron seabrum to wheat (Asker and Jerling 1992). However, success so far achieved has been limited. The molecular approach seems to be a better option for this task. One limitation to developing effective technology for such a transfer is lack of

sufficient basic information on the genetics of apomixis. Extensive studies are underway to identify and characterize the genes involved in meiosis and other events associated with apomixis in several species. These approaches have been elaborated by Maheshwari et al. (1998).

13 Transfer of Useful Genes to Crop Species Transfer of useful genes to crop species from other accessions or species through hybridiza­ tion has been the most effective approach in crop improvement programmes for many decades. Extensive breeding and selection have greatly reduced genetic variability of crop species. De­ pendence of agriculture on a narrow genetic base has markedly increased the vulnerability of present-day cultivars to a range of biotic and abiotic stresses. Although most cultivars have the genetic potential for high yield under opti­ mal agronomic practices (with high input of fer­ tilizers and other agrochemicals), their suscep­ tibility to biotic and abiotic stresses reduces yield significantly, often up to 50% (Boyer 1982). Broadening the genetic base of cultivated spe­ cies is an essential prebreeding requirement to maintain stability and further improve yield. Until recently crop improvement programmes concentrated almost exclusively on increasing yield and improving quality and genetic unifor­ mity. Remarkable success has been achieved over the years in fulfilling these limited objec­ tives. In recent years, however, the objectives of the breeding programme have become more diverse; some of the additional breeding objec­ tives apart from improvement in yield and qual­ ity are: 1. To develop varieties which are resistant/tol­ erant to biotic and abiotic stresses such as

diseases, pests, temperature, salinity and drought. 2. To develop varieties which have low require­ ment for chemical fertilizers, herbicides, pesticides, and other environment-unfriendly chemicals. Excessive use of agrochemicals has been a major environmental problem of modern agriculture. This aspect has so far been ignored by farmers as well as breed­ ers but has become increasingly important in recent years. The trend is to grow crops ‘organically’ without the use of pesticides, herbicides and chemical fertilizers. 3. To develop varieties that can be grown on marginal lands. A large area of agricultural land is being degraded by salinity, alkalinity, waterlogging, soil erosion and contamina­ tion with industrial chemicals. This is par­ ticularly severe in developing countries. As the recovery of these vast areas of land is not feasible because of the high cost in­ volved and/or lack of effective technologies, the alternative strategy of breeding crops that can be grown on such marginal lands has to be adapted. Realization of the aforesaid objectives is dif­ ficult and challenging. As many of the adaptive traits are controlled by multiple genes, transfer of such traits into a crop species by recombi­ nant DNA technology is not feasible; a

200 Pollen Biology and Biotechnology

conventional breeding programme through the sexual cycle has to be followed. Because of the constant erosion of genetic diversity, genes im­ parting the required traits are no longer avail­ able within the cultivars of crop species. How­ ever, a large number of wild and weedy rela­ tives of crop species form a good repository of such desirable traits (Harlan 1976). Unlike the early breeding programmes, which were largely confined to accessions within a species, presentday breeders have to extend the breeding programme across species limits, often to wild and weedy species (wide hybridization). A good number of examples of successful gene trans­ fer through wide hybridization are. already avail­ able (Table 13.1) (Hawkes 1977, Stalker 1980, Goodman et al. 1987, Hermsen 1992, Kalloo 1992, Chopra 1989, Chopra et al. 1996). Apart from its use in introgression of desir­ able genes into cultivars, wide hybridization is also effective in: (a) production of haploids through chromosome elimination (see Chapter 14); (b) development of new cytoplasmic male sterile (CMS) lines through synthesis of alloplasmic lines (see Chapter 5) and (c) syn­ thesis of new alloploid crops such as Triticale and Raphanobrassica (Sareen et al. 1992). The following are the major steps involved in transfer of genes, especially from wild relatives through sexual means:

1. identification of species/accessions having desirable genes, 2. production of hybrids between the cultivar and the parents with desirable gene(s) and raising of successive backcross genera­ tions, 3. screening a large number of plants in each backcross generation for identification of the required recombinants, and 4. stabilization of the recombinants. Figure 13.1 presents details of steps 2 and 3 using transfer of disease resistance as an ex­ ample. The conventional breeding programme implemented so far is often ineffective because of the presence of strong crossability barriers. Even in the absence of such barriers, conven­ tional breeding takes too much time. Integration of pollen biotechnological approaches into con­ ventional breeding greatly improves the effi­ ciency and reduces the cost and time of breed­ ing. Table 13.2 lists appropriate pollen biotech­ nological approaches, which can be integrated at different levels of the breeding programme.

USE OF POLLEN FOR IDENTIFYING PLANTS WITH DESIRABLE GENES A large number of accessions of related spe­ cies and often wild species of cultivars have to be screened to identify those with desirable

Table 13.1 Some selected examples of transfer of important traits from wild species to crop species through wide hybridiza­ tion (based on data presented in Goodman et al. 1987) Crop species

Donor species


Beta vulgaris Brassica napus Cucurbita pepo Gossypium hirsutum Lycopersicon esculentum Nicotiana tabacum Oryza sativa Solarium tuberosum Triticum aestivum Triticum aestivum

B. procumbens B. campestris C. landelliana G. raimonddii L peruvianum N. glutinosa 0. nivara S. demissum Aegilops comosa Secale cereale

Zea mays

Tripsacum dactyloids

Sugar-beet nematode resistance Club-root resistance Mildew resistance Rust resistance Nematode resistance TMV resistance Grassy stunt virus resistance Late blight resistance, leaf-roll resistance Stripe rust resistance Yellow rust resistance, leaf rust resistance, winter hardiness, stem rust resistance Northern corn leaf blight resistance

Transfer to Useful Genes to Crop Species 201 Table 13.2 Integration of pollen biotechnology into conventional breeding programme Major step/constraint Identification of accessions with desirable genes Spatial/temporal isolation of parental species Presence of pre- and post-fertilization barriers Multiplication of hybrids Hybrid sterility Handling of BC generations Stabilization of recombinants

Pollen biotechnological approach Use of pollen for screening Pollen storage Application of effective methods to overcome barriers Use of tissue culture technique Induction of amphiploidy Use of pollen for screening and application of selection pressure on pollen Induction of pollen embryos

genes before they can be transferred to the cul­ tivated species. Screening accessions for de­ sirable genes through the conventional method is laborious, time consuming and very expen­ sive. For example, screening of an accession for disease resistance using the conventional method involves growing plants under glass­ house conditions suitable for growth of the patho­ gen, spraying the plants with spores of the patho­ gen and scoring for the symptoms after allow­ ing sufficient time for pathogen establishment. It also restricts use of each plant for testing a single factor. Use of pollen for screening desirable genes offers many advantages; it is more effec­ tive and greatly reduces the time and cost of screening (Mulcahy 1984). As a single plant pro­ duces a large number of pollen grains, a range of factors can be tested against the pollen of the same plant. Use of pollen for screening for desirable genes is dependent on the expression of the required gene(s) in the pollen grain. Recent stud­ ies have conclusively established that a large number of genes are expressed in the pollen (Tanksley et al. 1981, Mascarenhas 1993, Hamilton and Mascarenhas 1997) and about 60% of them are also expressed in the sporophyte (Ottaviano and Mulcahy 1989) (see Chap­ ter 1). A simple method for testing the expres­ sion of gene(s) imparting resistance or tolerance to a particular stress in the pollen is to compare the responses of the pollen (Fig. 13.2) and the sporophyte under stress conditions. A number of investigations in several species have shown positive correlation between performance of the

pollen and that of the parent sporophyte under a range of stresses (Table 13.3), indicating that many of the adaptive genes are expressed in the pollen as well as the sporophyte. When such a correlation is established, plants resistant to a particular stress can be readily identified by studying the effect of stress on pollen germina­ tion and/or pollen tube growth under the stress condition (Hormaza and Herrero 1992, SariGorla and Frova 1997). For example, polleh grains from a plant resistant to a disease would be able to germinate in the presence of specific pathotoxin, while those from a susceptible plant would fail to germinate. Screening pollen for stress resistance/toler­ ance can be carried out using one of the follow­ ing four methods: i) in-vitro germination of pollen grains in the presence of stress, ii) pollination of stress-treated pistils with un­ treated pollen grains, iii) pollination of untreated pistils with pollen grains treated with stress, iv) application of stress to pollinated pistils. Although in-vitro germination is the most con­ venient and rapid method, it can only be used for those species in which satisfactory in-vitro pollen germination can be achieved. Pollen grains of many species, especially of 3-celled pollen species, are difficult to germinate in vitro (see Chapter 4). For such pollen other methods can be used. Pollen germination and pollen tube growth in the pistil can be conveniently studied through the use of aniline blue fluorescence (see Shivanna and Rangaswamy 1992).The number


Pollen Biology and Biotechnology





(good agronomic characters but disease susceptible) ■


F. HYBRID 1 (mixed agronomic characters and disease-resistant)




BC1 (selection of plants possessing as many good characters of the cultivar and disease-resistance of the wild species)



| I I +

bc2 (selection of plants as in BC1 and repetition of backcrossing until the desired recombinants [all good qualities of the cultivar and disease-resistance of the wild species] are realised )

Fig. 13.1 Protocol for transfer of desirable gene from wild species to cultivar through wide hybridization. Transfer of disease resistance is used as an example (after Shivanna and Sawhney 1997).

of pollen tubes growing in the style compared to the controls (untreated pollen and pistil) indicate the sensitivity of pollen grains to the stress con­ dition.

PRODUCTION OF HYBRIDS As pointed out earlier, strong crossability barri­ ers between parents which may operate before

Transfer to Useful Genes to Crop Species 203

100 -

_ 80-


l03e o Ç

'E o) 40 ■ c _ÇD

Ô °" 2 0 ­






Genotypes Fig. 13.2 Genotypic variation for low temperature tolerance using in-vitro pollen germination assay in Cicer arietinum. Differ­ ent accessions show marked variation in in-vitro germination response at 0°C which broadly reflects the responses of parents (based on K R Shivanna, N P Saxena and N Seetharama, unpublished). Table 13.3 Some of the agronomically important traits expressed in the pollen and the sporophytes (refer Sari-Gorla and Frova 1997 for additional references) Character



Tolerance to low temperature


Tolerance to salinity

Chickpea F1 hybrids of

Zamir et al. 1982, Zamir and Gadish 1987, Zamir and Vallejos 1983 Savithri et al. 1980 Sacheret al. 1983

Tolerance to heavy metals (zinc, copper) Tolerance to pathototoxins Tolerance to herbicides

Lycopersicum x Solanum Siiene dioica, Mimulus guttatus Maize to Helminthosporium Brassica to Alternaría Sugar-beet Maize

and/or after fertilization are the major constraints in any hybridization programme (see Chapter 10). Present-day plant breeders have a range of biotechnological methods available to over­ come such barriers. For selection of the most suitable technique, it would be necessary to identify the barrier.

Searcy and Mulcahy 1985a, b Laughnan and Gabay 1973 Shivanna and Sawhney 1993 Smith 1986 Sari-Gorla et al. 1989,1994

Effective Techniques to Overcome Physical Barriers The most important physical barrier for wide hybridization is non-synchronous flowering or geographic isolation of the parent species. A conventional breeding programme is ineffective

204 Pollen Biology and Biotechnology

in such crosses. Pollen storage is one of the simple and effective methods to overcome this temporal and spacial isolation of the parent spe­ cies.

Pollen storage Pollen grains can be stored for extended peri­ ods under suitable conditions to maintain their viability and used when required. Pollen storage has important applications in both basic and applied areas of reproductive biology. These are some of them: 1. Overcoming crossability barriers imposed due to temporal and spatial isolation of the parent species. 2. Eliminating the need to grow pollen parents continuously in breeding programmes. 3. Implementing supplementary pollinations to sustain yield (see Chapter 11). 4. Preserving genetic resources and provid­ ing germplasm for international exchange. 5. Ensuring availability of pollen throughout the year for studies on various aspects of pol­ len biology and pollen allergy. The need to establish ‘pollen banks’, which would ensure availability of pollen of the desired species at any time of the year and at any place, has long been emphasized. Pollen banks would greatly facilitate the breeding programme, par­ ticularly of tree species. Extensive studies have been carried out to assess various storage con­ ditions for extending pollen viability and pollen grains of a large number of species have been stored successfully (Table 13.4) (Knowlton 1992, Holman and Brubaker 1926, Visser 1955, Johri and Vasil 1961, King 1961, 1965; Stanley and Linskens 1974, Shivanna and Johri 1985, Towill 1985,1991, Hanna and Towill 1995, Barnabas and Kovacs 1997).The important methods used are briefly summarized below. Storage under low temperature (+4 to -20°C) and low humidity {